The "Process" of John Edward [Archive] - Page 2

voidx

12th August 2003, 10:04 AM

Originally posted by KelvinG
Regarding JE's show, I'm a little confused. I've heard believers on this board agree on many occasions that the only way to determine if JE is a real medium would be through valid scientific testing, and "Crossing Over" doesn't really prove anything (there is, after all nothing scientific about an edited program). And since JE has never been tested (and spare me any reference to the b.s. Schwartz "study"), we are all supposed to concede that no one can really sure whether he is real or not.

Yet, these same believers constantly point to his many "hits" on his TV show as proof as his legitimacy as a medium.

It seems to me you can't have it both ways. Either you reject his show as a source of proof (which strikes me as being the only logical approach) or you accept "Crossing Over" as a valid way to determine the legitimacy of JE as a real medium. In which case, your credibility as any kind of skeptic would be seriously compromised.

So, can we all agree, once and for all, that JE's show proves nothing.
Agreed. Some can take it that it "suggests" the "possibility" of something not mundane, however it proves nothing.

CFLarsen

12th August 2003, 10:14 AM

Originally posted by Leroy
I understand Instigator, but I think a better wording might have been "don't take Neo's word for it," instead of "Don't Listen to Neo." But, that is between you two, I am just an observer.

Perhaps, in time, you will actually contribute with something valuable to the discussions here, in a skeptical sense. Your whole output here seems to consist of nothing else but pointing out what you don't like about other posters, tell them how they should behave, and what they should think and say. You seem to have chosen the position as the Miss Manners of this board.

Just an observation.

TLN

12th August 2003, 10:49 AM

SteveGrenard

12th August 2003, 10:54 AM

KG: So, can we all agree, once and for all, that JE's show proves nothing.

Exactly. This is absolutely correct but there is negative bias from the so-called skeptic side also which keeps bringing up JE's performance on television as evidence of his cold reading, hot reading or guessing (warm reading). It is meaningless from both sides of the aisle.

BillHoyt

12th August 2003, 11:03 AM

Originally posted by SteveGrenard
Exactly. This is absolutely correct but there is negative bias from the so-called skeptic side also which keeps bringing up JE's performance on television as evidence of his cold reading, hot reading or guessing (warm reading). It is meaningless from both sides of the aisle.
Sorry, Steve, but no dice. It is simply not negative bias. It is, simply, Occam's razor at work. When I think there is nobody in my garage and yet I hear a sound from inside, I don't assume there is a unicorn in my garage. I go with more parsimonious hypotheses: maybe I am wrong that nobody is there. Maybe one of my dogs is out there. Maybe a racoon, a stray cat or possum. In my neck of the woods it could be a coyote, fox, woodcock, deer or bear. Maybe it isn't within the garage, but outside.

I don't assume poltergeists or spirits or that my dead grandfather-in-law who loved to tinker with cars is unable to let go of his lifelong passion.

It is, simply, Occam's razor.

Cheers,

voidx

12th August 2003, 11:59 AM

Thanz

12th August 2003, 01:56 PM

Originally posted by BillHoyt

At this point, I would call the analysis "suggestive".
"Suggestive" at best.

As I said, it is "suggestive". Moreover, it should be pursued further with forenames gleaned from U.S birth records (assuming a U.S. JE performance).
Don't you think forenames gleaned from U.S. death records would be far more relevant?

Not true. Depending on the analysis being done and the desired significance level, n=78 can be powerful in the statistical sense.
Maybe for some purposes, but I don't think that n=78 is adequate to tell us anything of real value for these purposes.

CFLarsen

12th August 2003, 02:01 PM

Originally posted by Thanz
"Suggestive" at best.

You are most welcome to come up with a better suggestion.

Originally posted by Thanz
Don't you think forenames gleaned from U.S. death records would be far more relevant?

Sure, if we can get them. Do you have them? No? Then perhaps we should work with what we got.

Originally posted by Thanz
Maybe for some purposes, but I don't think that n=78 is adequate to tell us anything of real value for these purposes.

You are most welcome to come up with better data.

renata

12th August 2003, 02:10 PM

Not quite what you need, but this is a list of names and their frequency from 1990 US Census- first and last, male and female.

http://www.census.gov/genealogy/names/

As to birth versus death records- I believe when JE throws out a name or an initial, he gets living relatives quite frequently as validation, so death records would not be necessarily the correct place.

CFLarsen

12th August 2003, 02:14 PM

Originally posted by renata
Not quite what you need, but this is a list of names and their frequency from 1990 US Census- first and last, male and female.

http://www.census.gov/genealogy/names/

This is the same reference I use. If anyone has earlier as well as more recent data, it would be most welcome.

Thanz

12th August 2003, 02:19 PM

Originally posted by CFLarsen
You are most welcome to come up with a better suggestion.
Suggestion for what? I think you are confused here. I am not saying that the method of comparing JE's guesses to some sort of letter name matrix is inherently bad, I am just saying that the results we have from the data at hand are "suggestive at best" that JE is using common name letters and therefore cold reading.

Sure, if we can get them. Do you have them? No? Then perhaps we should work with what we got.
I don't have birth records either. Do you? No? What do we have? A list of member names from some unknown organization. Hardly representative of the U.S. population, and hardly representative of the dead population either.

Sometimes, we just have to realize that the data we have is inadequate. That has been accepted (in this very thread) with respect to the show "Crossing Over". Due to editing, etc, we can't reliably use CO for this type of work.

So, I disagree that we should "work with what we got", as what we got is not adequate to give us any sort of accurate results.

You are most welcome to come up with better data.
I never thought that I wasn't. I just happen to have neither the time nor the inclination to come up with the better data. I leave that to people like you and your "Skeptical Report". But that doesn't mean that I am not entitled to point out problems in the data and analysis that I see here.

CFLarsen

12th August 2003, 02:33 PM

Originally posted by Thanz
Suggestion for what? I think you are confused here. I am not saying that the method of comparing JE's guesses to some sort of letter name matrix is inherently bad, I am just saying that the results we have from the data at hand are "suggestive at best" that JE is using common name letters and therefore cold reading.

Which is why I encouraged you to provide better data. Could you focus on that?

Originally posted by Thanz
I don't have birth records either. Do you? No? What do we have? A list of member names from some unknown organization. Hardly representative of the U.S. population, and hardly representative of the dead population either.

If you want to call the US Census Bureau an "unknown organization" then fine with me. You got anything better?

Originally posted by Thanz
Sometimes, we just have to realize that the data we have is inadequate. That has been accepted (in this very thread) with respect to the show "Crossing Over". Due to editing, etc, we can't reliably use CO for this type of work.

So, I disagree that we should "work with what we got", as what we got is not adequate to give us any sort of accurate results.

So, what? We step back and leave it, just like that? OMG, it's not "adequate", so let's leave it?? No, we need to work towards an answer.

Originally posted by Thanz
I never thought that I wasn't. I just happen to have neither the time nor the inclination to come up with the better data. I leave that to people like you and your "Skeptical Report". But that doesn't mean that I am not entitled to point out problems in the data and analysis that I see here.

You are quite right about that. However, you would just make a much more convincing case, if you could - just occasionally - bring something worthwhile to the table, instead of merely complaining about the works of others.

As it is, you do absolutely no work at all. You leave that to others, while you sit on the sideline, ready to point out possible flaws.

Thanz

12th August 2003, 02:59 PM

Originally posted by CFLarsen
If you want to call the US Census Bureau an "unknown organization" then fine with me. You got anything better?
I was referring to the source of juninho's numbers.

So, what? We step back and leave it, just like that? OMG, it's not "adequate", so let's leave it?? No, we need to work towards an answer.
I have no problem working towards an answer. I do have a problem with people thinking that results based on horribly flawed data have any meaning at all. Kerberos posted an analysis based on juninho's numbers, and was congratulated mightily for it. I, too, appreciate the work that he (?) did to come up with those numbers. However, to just accept them as meaningful without looking at the underlying data is sloppy. It was that reaction that I was primarily posting about.

You are quite right about that. However, you would just make a much more convincing case, if you could - just occasionally - bring something worthwhile to the table, instead of merely complaining about the works of others.

As it is, you do absolutely no work at all. You leave that to others, while you sit on the sideline, ready to point out possible flaws.
Well, yeah, it's a tough job - but somebody's got to do it. :p

Maybe you could just think of me as "peer review"?

But, you do make a point. If I have the time, I will try to bring more original analysis to the table for YOU to critisize. Fair?

Thanz

12th August 2003, 03:06 PM

Allright, I have actually found a fairly good source of names for this kind of analysis.

if you go here (http://www.ssa.gov/OACT/babynames/index.html) , you can find out the top 1000 names for the years 1990 to the present. You can also get the top 1000 names for each decade from from the 1900's through to 1990's. The source of the data is social secuity applications in the United States.

I don't have a lot of time to go through and count letters, however. Maybe if the task were broken down and spread across a number of posters it would be easier.

renata

12th August 2003, 03:17 PM

Originally posted by Thanz
Allright, I have actually found a fairly good source of names for this kind of analysis.

if you go here (http://www.ssa.gov/OACT/babynames/index.html) , you can find out the top 1000 names for the years 1990 to the present. You can also get the top 1000 names for each decade from from the 1900's through to 1990's. The source of the data is social secuity applications in the United States.

I don't have a lot of time to go through and count letters, however. Maybe if the task were broken down and spread across a number of posters it would be easier.

Good source! It is not that difficult, one can simply copy data into Excel and do a sort and a count. For example for 1900 names, there are 98 male A names, 52 B names, 71 C names, etc However, before that is done, exactly what do we want to count? Before we do the work, let's define it :)

Thanz

12th August 2003, 03:24 PM

Originally posted by renata

Good source! It is not that difficult, one can simply copy data into Excel and do a sort and a count. For example for 1900 names, there are 98 male A names, 52 B names, 71 C names, etc However, before that is done, exactly what do we want to count? Before we do the work, let's define it :)
Well, I think that a straight count is not the way you need to go - it needs to be weighted. For example, three of the top 5 men's names begin with "J".

So, I think that the number of people for each letter needs to be counted, and then we will know the actual distribution of letters across the sample.

renata

12th August 2003, 03:34 PM

Originally posted by Thanz

Well, I think that a straight count is not the way you need to go - it needs to be weighted. For example, three of the top 5 men's names begin with "J".

So, I think that the number of people for each letter needs to be counted, and then we will know the actual distribution of letters across the sample.

That is also pretty straighforward ( I love Excel)

So fo 1900 names,
There are 97 A names, for total of 31,584 names
51 B names for 7,734 names
71 C names, for 35,915 names
etc

Once the count is completed, what then?

(BTW I was one off on A & B names in prior post- my mistake)

neofight

12th August 2003, 03:38 PM

Originally posted by Instig8R

Neo, I did not just say "Don't listen to neofight about the editing". Did you only read the first line of my post??? I stated my opinion, and I explained why I disagreed with you.

LOL, 'g8R! Are you serious? Regardless of what you may have said afterwards, that does not alter the fact that you told
Kerberos quite directly NOT to listen to me about the editing!!! :rolleyes:

In other words, I challenged your opinion, and explained why I disagree with you over the editing... [/quote}

Had that been all you did, Instig8R, I would not have had a problem with it.

[quote]Just because someone disagrees with you is not justification for an ad hom... and making up accusations of hatreds and biases about other posters IS an ad hom. Are you trying to deflect attention from the Malibu Shrimp snafu? :D

As I said, it was not your disagreeing with me that I took exception to, it was your taking it upon yourself to instruct Kerberos to disregard my post, as though your opinion was worth considering, but mine was to be dismissed. That's all.

I was not attempting to deflect attention from anything, since I've often stated that I disagree with the significance you attach to those types of descrepencies that involve interpretation of symbols. We've gone over that many times. I know we disagree, and I accept that, no problem.

And I did not make up accusations of hatreds and biases about you. If you think *hate* is too strong a word to use, and perhaps you're right, it might be, then substitute *strong dislike* and/or *contempt*, and perhaps that would be more accurate.

With regard to the bias that I think you demonstrate against JE, there we will have to agree to disagree I think, because imo, your bias is not exactly understated. Come on now! I admit to my bias, 'g8R! Please don't deny yours. ;) .....neo

Thanz

12th August 2003, 03:42 PM

Renata - I am impressed by your excell skills. I would say that they were excel-lent, but that would be a heinous pun, so I won't.

This is the way I see it. If we combine the results from each decade, and then express each letter as a percentage of the total, we should have a decent approximation of the proportion of each letter for those born in the US in the last 100 years, which I think covers most (if not all) of the people that JE would be referencing. Then, for example, we can see if "F" names make up 7.8% of the total, how that relate to the number of times JE guess an "F" name.

The tougher part is getting enough JE guesses to make the comparison worthwhile.

Any other thoughts? Anyone? Bueller?

neofight

12th August 2003, 03:48 PM

[i]Have I seen someone who (a) I believe is not a medium and (b) who appears to do what JE does? Yes! The names Browne... [/B]

Well, Loki, like I said, I don't remember ever seeing Sylvia Browne actually do a mediumship reading, so I don't feel like I can compare her to JE.

In general, I don't remember too many people saying positive things about her, so I don't know how many of them would agree with you that she and JE have similar abilities.....neo

BillHoyt

12th August 2003, 04:02 PM

Originally posted by Thanz
"Suggestive" at best.
No, "suggestive" at least, not best. When scientists say "suggestive" they mean it lends support to an hypothesis. It is often used as an understatement.

Don't you think forenames gleaned from U.S. death records would be far more relevant?
JE's initials and names refer to people both living and dead. Death records would not be more relevant.

Maybe for some purposes, but I don't think that n=78 is adequate to tell us anything of real value for these purposes.
How about you get more specific, Thanz? Why is it inadequate here?

Cheers,

BillHoyt

12th August 2003, 04:10 PM

Originally posted by Thanz

Well, I think that a straight count is not the way you need to go - it needs to be weighted. For example, three of the top 5 men's names begin with "J".

So, I think that the number of people for each letter needs to be counted, and then we will know the actual distribution of letters across the sample.

You mean "normalized". For example, tally all the J names, divide by the total population, to obtain the control group's frequency.

Cheers,

neofight

12th August 2003, 04:52 PM

Originally posted by Instig8R
Neo, it is all opinion here, and I have never claimed otherwise. I should not have to preface every single sentence with "IMO", nor should I be bound by rules that you yourself do not live by.

Agreed. Just don't presume that your opinion trumps my opinion, and let other posters decide for themselves whether or not they want to ignore what I say. :)

Neo, this has come up a number of times. Whenever I criticize JE, I wind up being the target of an ad hom. You still seem to think it is justifiable to attack whoever criticizes JE. If we adhere to your rules, JE cannot be discussed, because anyone who utters a negative comment is attacked for doing so. This is a surefire way to stifle meaningful discussion.

Well, since we are only speculating about JE, I would have thought that it would be possible to discuss the merits, or lack thereof, of his claims to mediumship, without anyone constantly engaging in character assassination, but it seems I was mistaken. :confused:

It would be very helpful if you would supply me examples of the "nasty hateful rhetoric that I have used in discussing him". That is a very serious allegation on your part. I deny that I have ever engaged in any such behavior.

[b][quote]I do not use malicious language. Your allegation is downright defamatory. I would suggest that you either back-up your claim with evidence of the offensive rhetoric that you charge me with, or withdraw the claim.

Several times at TVTalk, I criticized JE and explained why I thought he was a fraud. However, instead of responding to my criticisms, you paraphrased my criticisms and referred to JE as "lowlife", etc., even though I did not engage in name-calling. Perhaps that is why you believe I was the author of ugly comments that YOU wrote about JE? ...Sort of like a poster's false memory syndrome.

If you look through the threads, I think you will see what I mean.

Well, if they were all in one thread where I could simply post the link, it would be easy, but they're not, Instig8R. Your derogatory comments are disbursed throughout a gazillion threads.

Friendship or not, these are impersonal debates on forums ... and detachment is the only way to debate an issue. The emotional state of a poster should not be presumed, because it is simply not relevant....

You're right, and I do try my best to keep the debates impersonal, but once in a while I drop the ball, and I let some remark of yours get to me, like the one you just made. I do admire your ability to completely detach yourself when you debate. You do an admirable job of that, for sure. ;)

Now, go find those offensive quotes of mine... I'm dying to read them! :)

A daunting task. I only searched through about two or three threads, but it took a lot more time than I'd like to devote to this task. I'll post what I did find, and hope that suffices. It's going to have to, because I'm just not doing this. It's really not all that important, and I'm only doing it because you wanted me to.

Post # 71 Instig8R 24.46.54.89 October 29th, 2002 08:15 PM

It seems to me that JE will say whatever suits his own selfish purposes at any given time.

As a believer, I understand that you are offended by what you perceive as undue criticism of JE, as someone you admire. As a disbeliever, I expect you to understand that I am offended by undue praise of JE, as someone I disdain.

Post # 165 Instig8R 24.46.54.89 November 26th, 2002 11:09 PM

He deflects the criticism away from himself, sometimes by undermining the intelligence of his fans, and more often by hiding behind widows and other bereaved individuals.

Post # 163 Instig8R 24.46.54.89 November 26th, 2002 09:47 PM

Hence, JE claims not to be concerned about critics/skeptics. The truth is that he is concerned about criticism, but is too cowardly to admit it. Therefore, he shifts the onus onto his fans, putting the insult and the injury on them. He then relies upon his fans to silence the critics. This is yet another reason why I consider him to be a manipulative s.o.b.

You want more than that, Instig8R, you'll have to find them yourself. You know they're out there. ;) As for me, I know for damned sure that I didn't imagine all the mean and defamatory posts that you made about JE, so if nothing else, I'm sure of what I said concerning your anti-JE bias........neo

neofight

12th August 2003, 04:58 PM

Originally posted by renata
Of course, above is a joke. But how you can pick an edited version of a show versus an unedited series of readings is unfathomable- and- in my mind against common sense.

Exactly, renata, and I'm so pleased to see that you understand this concept. You now understand precisely how John Hockenberry picked and chose exactly what he would air on his "Dateline" show so that he would make John Edward appear to have purposely cheated on national television. Thanks! You're a peach! :D ......neo

neofight

12th August 2003, 05:21 PM

Originally posted by Instig8R

P.S. I'd also like to know the identities of the other posters that you claim have commented on my "sometimes malicious language". Just who are these nameless posters?

As though knowing me, you really believe that I would state publicly what someone has told me privately. :rolleyes:

There were some unjustified personal attacks against me by atmytv and by sgrenard over at TVTalk. In response, other posters (like RC and Celter, for example) defended me because the attacks were extremely inappropriate and unjustified. (I believe atmytv apologized and we made up.)

Yes. Unjustified in your opinion, and as you said, in RC's and Celter's. Completely justified in others, including my own. True, atmytv is a gentleman, and he did apologize, but that does not mean that he felt he was mistaken, it just means that he was not about to make a big deal out of it. He's a nice man. This was what he initially said to you......followed by my response.......and then followed by his apology.

Post # 117 <atmytv> 64.133.202.146 November 12th, 2002 12:16 PM

I used to enjoy reading your posts. Even though they presented the other side, they were not cynical and distasteful the way they are now. You don't think you have changed, but you have. And, as I have acknowledged, I guess I have too, having more anger at the way John is treated and talked about here. It is one thing to have a discussion about whether such a thing as getting information off the internet is possible and the difficulties or ease of doing it and whether it explains all of his "hits". It is entirely different to state that John is a scum that clearly does it this way. Or to imply that anyone who does not recognize that is an idiot.

Post # 130 <neofight> 64.12.96.202 November 12th, 2002 10:20 PM

.....As far as atmytv's strong reaction to your sometimes rather caustic posts about JE, I do understand his upset. As you already know, I've experienced similar reactions to some of the things you have said about him, mainly because in my heart, I feel that he is a good person and that the attacks upon him are unwarranted. We who admire him and appreciate the work he is doing, do sometimes get offended at the level of disdain with which you and others regard him.

I know you don't get it when believers react this way, since you feel that you are directing your remarks at JE, and not at us, but I guess we just find it very unfair, and kind of like "dishonoring the living" as you put it. After all, if JE turns out to be the real deal, then your comments really will be looked upon as quite harsh and totally without merit.

And atmytv, Instig8R really does like to needle. ;) Seriously, after reading some of her posts, I feel a real need to see her in person or talk to her on the phone, so that I can remind myself what a nice person she really is, and why we've been friends for so long.

I think her evil alter-ego, Instig8R, is the one to blame for all this nasty stuff. ;) My friend does not even seem to be aware that she is coming off as nasty, as you can see from reading her response to you. She just enjoys the fracus. Try not to take her disparaging remarks to heart.
; ^ ) .....neo

(I should only take my own wise advise here, lol)

--------------------------------------------------------------------------------
Post # 136 <atmytv> 64.133.202.146 November 14th, 2002 07:53 AM

Instig8r. I apologize to you if anything I posted is taken as trying to diminish you in any way. That is not the intent. And I understand that you post your feelings about John and not about other posters.

My reaction to your posts comes from two sources for me, I believe. My memory might be faulty, but my impression was that when you and Dogwood first started posting, you both were clear that you had come away from seminars with disappointment about what you saw and serious questions about John. But your posts were not of the "John is scum" nature that they have been recently. So when I read your posts originally I never felt that I had to prepare myself for the nastiness toward John that was coming. So for me, it feels like you have become a cynic, rather than a skeptic. You are, of course, entitled to be whatever you like and to post whatever you want. As RC points out, he does not see it the same way, so I will accept the misperception as mine, and adjust to what in my mind is a new reality.

The second thing that has happened is, as Neo put it, that I am reacting to the incredible nastiness and disdain that has been directed towards John recently. It is very difficult to hear all of that stuff on a continual basis directed towards someone who you feel is real and doing a valuable service. In addition, the manner of the statements make it very difficult not to take it personally when the clear implication is that anyone who believes in this is a freaking dumbass bewildered gullible idiot. When you then add Cantata on top of that mix, it clearly got to me.

While you treat fellow posters with respect, perhaps you do not see how what you are saying about John can be interpreted as nasty and personal for those of us who just have a different viewpoint than yours. Now that I have vented my frustrations of the last several months, I will attempt to return to my thoughtful, rational, kinder self. Again accept my apologies if you felt unjustly attacked.

So I don't know about you, Instig8R, but I honored your request, and now I would just like to move on and start fresh, if that's all right with you.......neo

neofight

12th August 2003, 05:27 PM

Originally posted by Leroy

Yes, especially when he throws "suicide" in there.

Hi, Leroy! I just would like to clarify something here. He does not generalize here and just throw in "suicide", but only a drug overdose, accidental or otherwise, that would indicate a lot of "toxins" in the blood.

Anything like that would come under this symbol, along with the other problems that he mentions that could be detected in the blood. :) ......neo

neofight

12th August 2003, 05:32 PM

Originally posted by CFLarsen

Perhaps, in time, you will actually contribute with something valuable to the discussions here, in a skeptical sense. Your whole output here seems to consist of nothing else but pointing out what you don't like about other posters, tell them how they should behave, and what they should think and say. You seem to have chosen the position as the Miss Manners of this board.

LOL Claus. It's the right description, but it fits you, and not Leroy. How ironic is that? :roll:

Leroy has proven himself to be an objective as well as fair skeptic, Claus. You could stand to take a lesson! :rolleyes: ....neo

AlienX

12th August 2003, 05:35 PM

At last some progress and really Neo I thank you as I was beginning to think you were so closed minded and irrational as to not see such a basic principal of only looking at selective readings in isolation as being flawed.

Right so I assume the best thing to do is to look at a series of unedited readings not selected at all due to how good or how bad they are.. seems reasonable and you do say so yourself ;-)

Yet I suspect a few rolleyes coming on here as you point out that for whatever reason that looking at the multiple readings as a whole is somehow flawed..??

So far the only thread which does this is Renatas, it's irrelevent what the scoring and opinions are - we have multiple readings accurately transcripted - it's the best we have.

I'm glad you finally now realise and say so yourself just as holding up specific bad readings is just as bad as pointing the specific good readings.

Now perhaps you could explain this basic principal to JE and give us access to unedited tapings of crossing over as i'm sure due to your comments you agree that this edited show is missrepresenting what actually happens.

Thats just great so I suppose we will see an end to threads which contain a single specific hit reading... at last I agree with you 100%...... progress ;-)

AlienX

neofight

12th August 2003, 05:55 PM

Originally posted by AlienX
At last some progress and really Neo I thank you as I was beginning to think you were so closed minded and irrational as to not see such a basic principal of only looking at selective readings in isolation as being flawed.

Don't mention it, AlienX. ;)

Right so I assume the best thing to do is to look at a series of unedited readings not selected at all due to how good or how bad they are.. seems reasonable and you do say so yourself ;-)

Yet I suspect a few rolleyes coming on here as you point out that for whatever reason that looking at the multiple readings as a whole is somehow flawed..??

So far the only thread which does this is Renatas, it's irrelevent what the scoring and opinions are - we have multiple readings accurately transcripted - it's the best we have.

LOL, no, I'll spare you the rolleyes, AlienX. I realize that they are the only unedited readings that we have, but the reason that I don't believe it's fair to use those readings is because no matter what some might say, those of us who are familiar with JE's readings, from watching the show, and/or from going to seminars, know for a fact that these short little phone readings are not representative of a full JE reading. They're just not.

An average JE reading goes on for quite a while. These mini-readings are timed in mere seconds. They can't measure up to the usual length reading. It's impossible, and because of that fact, and because there is no immediate follow-up where the sitter can verify some of the things that JE said that flew by too fast for them to react to initially, these readings are less than ideal to work with. Far less than ideal......neo

edited to chance the word "edited" to "unedited"

BillHoyt

12th August 2003, 06:31 PM

Originally posted by neofight
As though knowing me, you really believe that I would state publicly what someone has told me privately. :rolleyes:
What duplicity! You use private messages to make a claim and then hide behind their private nature to refuse to back up that claim.

Stop wasting our time and JREF bandwith!

BillHoyt

12th August 2003, 06:36 PM

Originally posted by neofight
LOL, no, I'll spare you the rolleyes, AlienX. I realize that they are the only edited readings that we have, but the reason that I don't believe it's fair to use those readings is because no matter what some might say, those of us who are familiar with JE's readings, from watching the show, and/or from going to seminars, know for a fact that these short little phone readings are not representative of a full JE reading. They're just not.

More duplicity! Unedited readings aren't representative because you say so. No other evidence. Somehow we are to believe the edited TV shows are more representative. What a lovely veil this broadbrush makes. We will never know the truth of JE with your system, will we? We can only trust readings edited by JE and his producers. Wow.

neofight

12th August 2003, 06:46 PM

BillHoyt, you are a most unpleasant person. :p ......neo

neofight

12th August 2003, 06:54 PM

Originally posted by BillHoyt

More duplicity! Unedited readings aren't representative because you say so. No other evidence. Somehow we are to believe the edited TV shows are more representative. What a lovely veil this broadbrush makes. We will never know the truth of JE with your system, will we? We can only trust readings edited by JE and his producers. Wow.

The readings that are allowed to go on until they stop are the more representative readings, Hoyt, regardless of whether they are edited or not, as in the gallery, OR at a seminar.

Why don't you make the attempt to go to one of JE's seminars so you can judge for yourself what a real unedited full-length JE reading looks like, instead of bitching and moaning about the issue?

We can take up a collection if you don't want to put money in JE's pocket. I'll even chip in, if my share doesn't come to more than $10. :rolleyes: .......neo

Loki

12th August 2003, 06:58 PM

neofight,

those of us who are familiar with JE's readings, from watching the show, and/or from going to seminars, know for a fact that these short little phone readings are not representative of a full JE reading. They're just not.
...
these readings are less than ideal to work with. Far less than ideal
I understand why *you* think this - but why doesn't JE say anything like this? Why does he do "far from ideal" and "not representative" readings? Why doesn't he preface these readings by saying something like "I'd just like to say before we start that these readings will be far from ideal, because of the restricted time/format.". You say he can't perform well, and these readings shouldn't be seen as "representative". I haven't heard him say that at all. You know more about this than he does?

neofight

12th August 2003, 07:05 PM

Originally posted by Loki
neofight,

I understand why *you* think this - but why doesn't JE say anything like this? Why does he do "far from ideal" and "not representative" readings? Why doesn't he preface these readings by saying something like "I'd just like to say before we start that these readings will be far from ideal, because of the restricted time/format.". You say he can't perform well, and these readings shouldn't be seen as "representative". I haven't heard him say that at all. You know more about this than he does?

Loki, I'm certain if he was asked the question, he would have to agree that having the people he is reading right there in front of him has to be less confusing to him than having ten people holding on phone lines.

I mean, it's common sense. At least when they are in front of him physically, he feels a "pull" as to whom he is supposed to be with that leads hiim to the right area. I doubt that happens on the phone......neo

SteveGrenard

12th August 2003, 07:09 PM

Each call is about 1 minute or less. The intent of LKL is to have him take as many of these 1-minute calls as possible. JE doesn't say 1-minute is not enough time and these calls a bunch of crap because King won't let him say it although I have sensed his discomfort in this regard. He's on LKL for international exposure. His promoters got him on to sell books or whatever and have his image sent out worldwide. LKL shows at 2 or 3 AM in the UK and a lot of people stay up to watch him although I don't know why.

Whoever, said JE has control over the calls is mistaken. King has a switch. When the call goes over, even if the caller or JE is not finished., he (King himself) cuts them off. Next caller. Cut. Next caller. Cut. Next caller. Cut.

In short, this call in stuff on LKL is a bunch of crap. It can't be used to rate anything for or against JE. And no TV performance of any kind can be the scientific basis for determining a medium's validity which is why the bandwith wasted on this is aburd.

neofight

12th August 2003, 07:12 PM

Loki, I forgot to respond to the rest of your question. You're asking why he doesn't give a disclaimer of sorts, before he begins these phone readings?

I don't know why, but I suppose it's possible that he doesn't want to appear like he is making excuses for poor performance even before he begins. As it is I've heard him criticized by some skeptics for trying to "lower expectations" when he warns a sitter that the person they most want to hear from might not come through. (although most of the time, they do)

And I'm not sure that he would do these tv guest appearances at all, except that perhaps it is required of him contractually every time his publisher releases a new book.......neo

KelvinG

12th August 2003, 07:23 PM

Originally posted by SteveGrenard
KG: So, can we all agree, once and for all, that JE's show proves nothing.

Exactly. This is absolutely correct but there is negative bias from the so-called skeptic side also which keeps bringing up JE's performance on television as evidence of his cold reading, hot reading or guessing (warm reading). It is meaningless from both sides of the aisle.

So, if we can't trust JE's readings on LKL, or his readings on CO, then why is there even discussion about him.
Until some legitimate scientific testing is done, we are left with no answers.
And because of that, is it not more logical to conclude that JE's technique is nothing more than cold reading?
That doesn't mean that the possibility of him being a medium can be discounted entirely, but it does mean a little common sense makes the medium scenario incredibly far fetched.

If we can only standby and say "JE might be a medium or he might not be a medium, we just don't know for sure", then the believers lose.
Why do they lose? Because the burden of proof is on them. It's not up to the skeptics to prove JE is not a medium (although they certainly seem to propose the vastly superior arguments), it's up to JE to prove he is a medium.

Until he does, I have to assume, through common sense and reason, that he is not a medium, but just a talented cold reader.

If anyone thinks otherwise, prove me wrong. Of course, you can't. And don't ask me to prove that he truly isn't a medium. The burden is not on me. Why? Because the last time I checked the scientific community has not embraced Edward as the amazing phenomenon that he claims to be. Science textbooks have not been rewritten to include his powers which would defy known laws of science.
Until that happens, JE is simply a novelty act, a gimmick if you will.

Sorry, but as things currently stand John Edward is going to be remembered as an entertainer in the same category as David Copperfield. (except not nearly as famous or talented as Copperfield).
Kind of sad for someone who supposedly talks to the dead!!

renata

12th August 2003, 07:42 PM

Originally posted by neofight

Exactly, renata, and I'm so pleased to see that you understand this concept. You now understand precisely how John Hockenberry picked and chose exactly what he would air on his "Dateline" show so that he would make John Edward appear to have purposely cheated on national television. Thanks! You're a peach! :D ......neo

I do of course understand it-I brought the issue up and have been doing so for a while. :i: You, however, appear to take my point about unreliability of the evidence you rely on and attempt to divert the topic elsewhere.
You continue to use the edited CO and your unreliable memory of the seminar readings (later edited, as proven by Instigater) as evidence of his abilities. You continue to discard LKL readings as useless, snippets, unacceptable. However, when Clancie quoted the Timothy reading from LKL, probably the best of all the readings he did there http://www.randi.org/vbulletin/showthread.php?s=&threadid=23120 you did not object that it was a snippet! You said
Good choice, Clancie! And this sitter is extremely tight-lipped!

So your objections are only when totality of evidence is presented and examined- you do not mind a single decent reading taken out of context of many abysmal ones. :bs:

That smacks of intellectual dishonesty and self deception, Neo.

Edited to test out the new irony and bs meters :)

renata

12th August 2003, 07:47 PM

Originally posted by Thanz
Renata - I am impressed by your excell skills. I would say that they were excel-lent, but that would be a heinous pun, so I won't.

This is the way I see it. If we combine the results from each decade, and then express each letter as a percentage of the total, we should have a decent approximation of the proportion of each letter for those born in the US in the last 100 years, which I think covers most (if not all) of the people that JE would be referencing. Then, for example, we can see if "F" names make up 7.8% of the total, how that relate to the number of times JE guess an "F" name.

The tougher part is getting enough JE guesses to make the comparison worthwhile.

Any other thoughts? Anyone? Bueller?

*groan* :) I hoped you would resist! Or at least make a better pun! ;)

I can add all the names up- but if we want to know how many types of names, it will be more difficult, because there would be no exact overlap necessarily between 1900 and 1970, and I don't feel like calculating manually. But perhaps only the numbers of names in total for those years are relevant. In that case, I can do the math relatively easily- but I do not want to do it unless someone somewhere will use it for analysis! :)

And whereas we have this name data, it will be difficult to gather the untainted transcripts for analysis.

SteveGrenard

12th August 2003, 07:51 PM

KG: So, if we can't trust JE's readings on LKL, or his readings on CO, then why is there even discussion about him.
Until some legitimate scientific testing is done, we are left with no answers.

Reply: I AGREE.

KG: And because of that, is it not more logical to conclude that JE's technique is nothing more than cold reading?

Reply: No, because if you a true skeptic who places your faith in the scientific method you cannot make any conclusions on him until he has b een scientifically debunked or validated.

KG: That doesn't mean that the possibility of him being a medium can be discounted entirely, but it does mean a little common sense makes the medium scenario incredibly far fetched.

Reply: It doesn't mean anything actually. And "far fetched" is hyperbole and not a scientific conclusion.

KG: If we can only standby and say "JE might be a medium or he might not be a medium, we just don't know for sure", then the believers lose.

Reply: There can be no winners or losers. This is not a contest or a race. With scientific evidence there is no winners and no losers.
The truth wins. Nothing else really matters.

I agree. He might or he might not. But do we know? NO.

KG: Why do they lose? Because the burden of proof is on them. It's not up to the skeptics to prove JE is not a medium (although they certainly seem to propose the vastly superior arguments), it's up to JE to prove he is a medium.

Reply: JE doesn't seem to care what you or I think and doesn't consider it up to him to prove anything. However other mediums are interested in research and are participating in research and, over the years past, have done so as well. There are three serious books that discuss these: Braude: Immortal Remains; Gauld: Mediumship-A Century of Investigations and the original one that started it all, by FWH Myer's. Neither believers or non-believers have any statutary burden placed on them. Individuals can examine the evidence and decide for themselves.

KG: Until he does, I have to assume, through common sense and reason, that he is not a medium, but just a talented cold reader.

Reply: Assumptions are not scientific and the word doesn't appear in the scientific method.

KG: If anyone thinks otherwise, prove me wrong. Of course, you can't. And don't ask me to prove that he truly isn't a medium. The burden is not on me. Why? Because the last time I checked the scientific community has not embraced Edward as the amazing phenomenon that he claims to be. Science textbooks have not been rewritten to include his powers which would defy known laws of science.

Reply: You are incorrect. Edward has neither been asked nor has he denied any requests to be tested further by a scientific
researcher and institution. There is nothing amazing about JE and I would ask you to show me where he claims he is amazing. Randi claims he is amazing "The Amazing Randi." I have never seen JE do that. There are plenty of people who claim to do what JE does and some of them claim to do it a lot better than he does. See the above references. JE is not mentioned in any of them.

KG: Until that happens, JE is simply a novelty act, a gimmick if you will.

Reply: I am not sure what a gimmick or novelty act is in this context but if this is how you want to characterize his performance you are certainly entitled to that opinion. We are all critics.

Loki

12th August 2003, 08:11 PM

neofight,

...I'm certain if he was asked the question, he would have to agree that having the people he is reading right there in front of him has to be less confusing to him than having ten people holding on phone lines.

I mean, it's common sense...
But what has commonsense got to do with this? If commonsense was applicable, then surely JE would be able to get letters (even words) from the spirits - they are, after all, the simplest and most common 'symbols' within the 'frame of referecen' of JE, the spirit, and the sitter. But JE tells us "it doesn't work like that", so obviously "commonsense" is not a useful tool in examining JE's process. If you can't use commonsense, and JE has not offered any complaint or 'disclaimer' about LKL style readings, then on what basis do you conclude they are "not representative"?? Isn't it true that your main reason for feeling this readings are "not representative" is because they generate far fewer hits??

We both start from "LKL readings have far fewer hits".

You conclude that "LKL readings are 'process unfriendly'".

I can't see how you reach that conclusion, without just assuming it. 'Commonsense' can't get you there (with any degree of reliability), and JE doesn't tell you this. So how?

Actually, I think you do have one point that we could agree on that makes LKL readings less that ideal....
..and because there is no immediate follow-up where the sitter can verify some of the things that JE said that flew by too fast for them to react to initially,...
Given the speed of the reading, the sitter has to respond instantly with hit/miss. This increases the chances for error, where the sitter might fail to find a hit, or incorrectly validate information that doesn't actually fit. The shorter the reading, the more likely these sorts of errors will occur. Of course, it's possible that these errors might work either way, so it's hard to conclude that this "lack of considered validation" by the sitter is responsible for the poorer reading.

KelvinG

12th August 2003, 08:20 PM

SG: No, because if you a true skeptic who places your faith in the scientific method you cannot make any conclusions on him until he has b een scientifically debunked or validated.

OK, fair enough. Perhaps I was jumping to conclusions. So, I can assume that you are absolutely in the middle on JE. You have no more belief one way or the other whether JE is real or not?

SG: It doesn't mean anything actually. And "far fetched" is hyperbole and not a scientific conclusion.

Yes, but it sounds cool.:D But fair enough again, the word is one I use of my own free will, perhaps incorrectly.

SG: There can be no winners or losers. This is not a contest or a race. With scientific evidence there is no winners and no losers.

Wow, considering how heated debates get on this forum and the amount of name calling that occurs I concluded otherwise.

SG: JE doesn't seem to care what you or I think and doesn't consider it up to him to prove anything. However other mediums are interested in research and are participating in research and, over the years past, have done so as well. There are three serious books that discuss these: Braude: Immortal Remains; Gauld: Mediumship-A Century of Investigations and the original one that started it all, by FWH Myer's. Neither believers or non-believers have any statutary burden placed on them. Individuals can examine the evidence and decide for themselves.

Like I said in my earlier post, I will wait until mainstream science embraces these studies and declares that paranormal phenomenon actually do exist. As a non-scientist and somehow who only follows science casually, I have to defer to the opinions of experts. If I read the data from the above mentioned studies, I probably wouldn't be able to interpret it. I will rely on the majority opinion in the scientific community. And so far, that doesn't include paranormal phenomenon.

SG: JE doesn't seem to care what you or I think and doesn't consider it up to him to prove anything.

Most con men won't really care what their critics think. And yes, I called him a con man. I realize it might be unscientific of me, but I'm just going by a hunch on this one. Hey, who knows, maybe this hunch is a paranormal experience! We'll never really know without scientific testing.;)

SG: You are incorrect. Edward has neither been asked nor has he denied any requests to be tested further by a scientific

I think this goes to show exactly how seriously his so-called "abilities" are taken. Someone who can talk to the dead and the scientific community isn't interested!! Mind boggling!!

SG:There is nothing amazing about JE and I would ask you to show me where he claims he is amazing.

Yes, I agree, there is nothing particularly amazing about cold reading. However, if you think he might actually be talking to the dead, then holy sh*t, I do think that is amazing. I think that is beyond amazing.
And I never said that Edward claimed he was amazing.

SG: I am not sure what a gimmick or novelty act is in this context but if this is how you want to characterize his performance you are certainly entitled to that opinion. We are all critics.

Fair enough again. I do characterize him this way. This simply falls into the "call em as I see em category."
Granted, it ain't scientific, but I'm not always scientific either.

Thanks Steve.

Loki

12th August 2003, 08:35 PM

SteveGrenard,

(KelvinG wrote) : And because of that, is it not more logical to conclude that JE's technique is nothing more than cold reading?

(SteveGrenard wrote) : No, because if you a true skeptic who places your faith in the scientific method you cannot make any conclusions on him until he has b een scientifically debunked or validated.
Nonsense! An "open mind" doesn't not mean the same thing as "cannot make a conclusion". It's *possible* that dogs and cats are engaged in a war to conquer the Earth, and are just using humans as resource gatherers. There is no contradiction in saying "I am skeptical of this" and "I believe it to be false". Skeptical means "prepared to base my conclusions on the available evidence".

What is the available evidence in favor of JE? As you agree, none.

What is the available evidence against JE? At the very minimum, the existing body of knowledge regarding physics, and the proven track record of humans to lie and cheat.

Is this *conclusive* evidence against JE? I think so! He offers absolutely nothing in support of his claim, and needs to overturn the scientific basis of modern knowledge to establish this claim. He doesn't have to be scientifically debunked in order for me to say "fake" - science already says he can't do what he claims. He does have to be scientifically validated for me to say "he's genuine".

. However other mediums are interested in research and are participating in research and, over the years past, have done so as well.
Slight change of tack here (by me) - why don't mediums do their own research? You seem to think that JE (or anyone else) needs to be involved in institutional research? Sure, that's wereh it has to end up - but why are "beleiver" groups (say, you, your wife, and your NY trance medium) interested in generating some data to present to the public? Why doesn't this *ever* happen :

Steve : "Gee Trance Medium, you're 100% accurate! I'd like to propose a few tests, to gather some data that will change the human race!"
TranceMedium : "Sure! Lets go for it - what do you suggest?"
Steve : "Well, there are a few simple things we can do - how about next week I arrange a simple blind test of a few friends of mine - you can read them multiple time from behind a curtain, and see if you can get the same readings?"
TranceMedium : "Yes, that sounds interesting - I'd like to see how far my powers extend. Let's do it!"

Now, *if* that ever happened it would prove nothing of course. Replication is the key. But why doesn't the "medium industry" *ever* produce this sort of data? JE probably has enough wealth to establish the "John Edward Psychic Research Institute" if he felt the desire. Why this total lack of interest in any self-organised structured testing at all? Doesn't it say something?

SteveGrenard

12th August 2003, 08:35 PM

KG: OK, fair enough. Perhaps I was jumping to conclusions. So, I can assume that you are absolutely in the middle on JE. You have no more belief one way or the other whether JE is real or not?

Reply: Yes this correct. Wihout the requsiite testing, where JE as an individual is concerned, a true skeptic can never be certain. All we have to date on this personage is the Univ of Az testing which has been critricized and deemed flawed in someway by cynics and skeptical reviewers. Thus where JE is concerned more testing would be required since the Az testing did not find him to be a fake but it did not satisfy the cynics and skeptics that he was real either. Based on the skeptical critiques and the failure to get a straight answer as to what would be acceptable and 10 different opinions for a protocol including some which are pretty ridiculous (such as Randi's Sylvia Browne design) it may be impossible to get any cynic/skeptic's nod on any protocol in advance that would satisfy such a person. At
that point researchers just have to go forward anyway and do so with the best design and advice they can muster.

KG: Wow, considering how heated debates get on this forum and the amount of name calling that occurs I concluded otherwise.

Reply: There are cynics here who are rhetorical skeptics; others who try to counter honest debate and denigrate those with whom they do not agree by ad hominem attacks. If you have nothing left to say try "liar" is one frequent technique. It is designed to hurt the credibility of ones opponents. It is now so commonly used by certain parties it is hardly worth mentioning anymore and has lost all of its true meaning. Such folks get what they give and should expect no less. There is no turning the other cheek here when attacked in this way.

SteveGrenard

12th August 2003, 08:56 PM

--------------------------------------------------------------------------------L:Nonsense! An "open mind" doesn't not mean the same thing as "cannot make a conclusion". It's *possible* that dogs and cats are engaged in a war to conquer the Earth, and are just using humans as resource gatherers. There is no contradiction in saying "I am skeptical of this" and "I believe it to be false". Skeptical means "prepared to base my conclusions on the available evidence".

Reply: You can make all the "isn't it possible arguments" in the world, it still is not scientific. I never used the term open mind and unless JE is scientifically debunked or discredited or validated, I can't be certain about him.

L: What is the available evidence in favor of JE? As you agree, none.

Reply: We are NOT talking about available evidence. We are talking about scientifically obtained evidence under controlled conditions. The available evidence is crapola. Any pro or con evidence you get off the television or from uncontrolled public performances or 60 second telephone readings on LKL is worthless. I agree: none for, and none against.

L: What is the available evidence against JE? At the very minimum, the existing body of knowledge regarding physics, and the proven track record of humans to lie and cheat.

Reply: Ah, the liar and cheat argument. Isn't it possible? I dont buy the isn't it possible argument no more than I buy the "isn't it impossible" either. These are meaningless terms. With respect to the physics evidence, there is enough we don't know to make a similar conclusion ....and in the world of quantum physics, the isn't possible or its not impossible arguments run rampant on a daily basis and are used by physicists to advance their understanding of such possibilities. But I still don't buy them.

L: Is this *conclusive* evidence against JE? I think so!

Reply: I think not. It neither supports or excorciates him.

L: He offers absolutely nothing in support of his claim, and needs to overturn the scientific basis of modern knowledge to establish this claim. He doesn't have to be scientifically debunked in order for me to say "fake" - science already says he can't do what he claims. He does have to be scientifically validated for me to say "he's genuine".

Reply: After you have read Braude, Myers and Gauld and many of the supporting references in there , come back and tell me that "science" whoever he is or whatever that is as an entity,
says he can't do what he claims. The science in these studies say otherwise. More recent studies by Robinson and Roy in the UK, the work done by Honorton, Bem, Parker et al on telepathy, all this adds up to the conclusion that science says nothing of the kind that you claim it says. This doesn't mean JE is real but you can't label him as a fake based on this argument. In the face of other evidence, I couldn't care less about JE or his claim but I still require him to be validated or discredited under controlled conditions by (a) legitimate scientific researcher(s) on an academic institutional evel.

I can understand why you are now turning away from the scientific method when it is being invoked against your worldview. It is fascinating how cynics use that argument when it suits them and disdain it when it doesn't. It is too funny.

L: Slight change of tack here (by me) - why don't mediums do their own research? You seem to think that JE (or anyone else) needs to be involved in institutional research? Sure, that's wereh it has to end up - but why are "beleiver" groups (say, you, your wife, and your NY trance medium) interested in generating some data to present to the public? Why doesn't this *ever* happen :

Reply: I think there would be immediate outcries from people such as yourself citing a conflict of interest if JE funded his own research. I have not asked the NY trance or the British trance if they are interested and dont intend to. I have no personal contact w/them and my wife and I saw them anonymously and professionally. We were not doing research on them. The British medium is already one of many who works with the SPR (I was advised of this afterwards by Keen who knew him from that work) and the NY lady is elderly and keeps to herself, makes no claims, wants no publciity and wouldn't know anything about research if it jumped up in her face. She is not very rich, lives in a tiny house in a middle class neighborhood. You can pick on her if you want but I find it ridiculous to do so with so many hundreds, probably tousands of practicing mediums worldwide, many of who are participating in research. The 40 or so mediums working with Roy and Robinson in Edinburgh are an example.

So don't worry, Research is occurring, has been published in the past and will be in the future. Because you do not want to acquaint yourself with this data doesn't mean it doesn't exist. I
am constanty amazed at how naive and impoverished such arguments truly are but then again I am not. Until I took the time to research this I was similarly inclined so can hardly blame you.

BillHoyt

12th August 2003, 09:13 PM

Steve,

When the h*** are you going to start backing up your outlandish claims about Einstein? About Occam's razor? When are you going to stand up and face the facts. This last tirade about skeptics abandoning science when convenient is entirely based on your notion that Occam's razor is no longer used in science.

Marshall the evidence or stop it. Those are the rules here. Choose your path:

o Support your claim with evidence
o State you believe it despite having no evidence
o State you refuse to answer the question

Loki

12th August 2003, 09:23 PM

SteveGrenard,

...unless JE is scientifically debunked or discredited or validated, I can't be certain about him.
That's your *personal* choice Steve - to weight the the evidence and conclude "I'm not sure". But what I'm replying to is what you said : "if you a true skeptic who places your faith in the scientific method you cannot make any conclusions on him until he has b een scientifically debunked or validated".

Again, this is not so. There *is* evidence, and it's entirely possible to be a skeptical and to say *right now* that JE is a fake.

Any pro or con evidence you get off the television or from uncontrolled public performances or 60 second telephone readings on LKL is worthless. I agree: none for, and none against
Yes, the value of CO., LKL, etc evidence is zero one way of the other - but that's not *all* the evidence available!

With respect to the physics evidence, there is enough we don't know to make a similar conclusion ...
Ah ... "Mediumship of the gaps". What we do know says it can't happen. You think the answer lies in what we don't. Okay - but don't try to call that "following the evidence".

"What we don't know" is precisely that Steve - something we don't know.

So he must be scientifically debunked.
Nope. If I introduce you to a dowser tomorrow is your initial reaction :

a. I think he's genuine
b. I think he's a fraud.
c. I don't have enough data to decide one way or the other.

Seems like you'd choose (c)? Sorry, but *until* the dowser establishes a positive result, he doesn't deserve option (c) - because he's working against the established laws of physics.

I can understand why you are now turning away from the scientific method when it is being invoked against your worldview. It is fascinating how cynics use that argument when it suits them and disdain it when it doesn't. It is very funny.
It's at moments like this that I understand why so many people have strong issues with you. Could you say something less relevant and more insulting? Probably, but this is a pretty good effort.

I think there would be immediate outcries from people such as yourself citing a conflict of interest if they funded their own research.
You somehow missed my point. Here, I'll type a little slower for you - why isn't this being done? You seem to say "because critics will claim 'not valid'" Of course they will! But why do the believers not want to do this? Just fear?

I have not asked the NY trance or the British trance if they are inetrested. ...We were not doing research on them.
(Sigh) Yes, Steve, I can read - i know you aren't doing this. What I wondering is why *no one* is doing this. You answer appears to be "I don't know"?

So don't worry, Research is occurring, has been published in the past and will be in the future.
Yep. I await something interesting. Something beyond "statistical anomalies discovered through meta-analysis" would be good. Oh wait, you found a trance medium that is 100% accurate on almost 200 different items of information! How about we test that? Oh, sorry she's "too old" and "not interested". Okay.

Because you do not weant to acquaint yourself with this data doesn't mean it doesn't exist.
Just because you want to accept anything and everything placed in front of you doesn't mean it does exist.

I am constanty amazed at how naive such arguments truly are but then again I am not. Until I took the time to research this I was similarly inclined so can hardly blame you.
Patronising little *****. "I used to be dumb like you, until I saw the light". Well, I'm convinced!

SteveGrenard

12th August 2003, 10:21 PM

L: Again, this is not so. There *is* evidence, and it's entirely possible to be a skeptical and to say *right now* that JE is a fake.

Reply: Aside from evidence gleaned from uncontrolled public performances of Edward, what other evidence is available that you cite? Please provide this.

L: Yes, the value of CO., LKL, etc evidence is zero one way of the other - but that's not *all* the evidence available!

Reply: Yes, what is the other evidence? Please don't tell me its mystery-man O'Neill. That was the biggest piece of hoaxing b.s. randi has come up since Project Alpha. Of course randi has warned us he is a trickster, charlatan and a liar so we
shoudnt be surprised if Mr O'Neill was more than a wee bit coached on this account of his.

L: Ah ... "Mediumship of the gaps". What we do know says it can't happen. You think the answer lies in what we don't. Okay - but don't try to call that "following the evidence".

"What we don't know" is precisely that Steve - something we don't know.

Reply: From my POV more like an argument from crippling complexity.

L: Nope. If I introduce you to a dowser tomorrow is your initial reaction :

a. I think he's genuine
b. I think he's a fraud.
c. I don't have enough data to decide one way or the other.

Reply: I cannot even begin to answer this question except to say I have no information, knowledge nor have I studied or been shown any evidence about dowsing so I cannot answer the question. I guess I get a zero on that one. If you wish to compare apples and oranges, try doing it with someone else.
I can taste the difference.

The etablished laws of physics are what? Can you summarize them for us? Is quantum theory turning physics on its rear end or not? I think there are a lot of physicsts who feel that it is.
List ALL the "established" laws of physics and what establishes them.

L: It's at moments like this that I understand why so many people have strong issues with you. Could you say something less relevant and more insulting? Probably, but this is a pretty good effort.

Reply: Its a simple and truthful from the heart observation. I have Claus Larsen yelling and screaming about the scientific method, I have Hoyt yelling and screaming about Occam's Razor and the scientific method, we are asked for falsifiable hypotheses and then, to suit your own thesis, you don't need of any of this. You can't have it both ways.

L: You somehow missed my point. Here, I'll type a little slower for you - why isn't this being done? You seem to say "because critics will claim 'not valid'" Of course they will! But why do the believers not want to do this? Just fear?

Reply: You don't know if its being done and never will. If JE or any other wealthy medium is funding mediumship research on mediums other than themselves, you will no more hear of it just as we do not know who gave Randi his million dollar prize money.
You feel that it is okay for Randi to claim he has a million dollar prize but not reveal who the donor is but you want research that is occuring right now to be funded in the open by donors who are made known to you? You can't have it both ways. JE will also never fund research to vaidate himself, of that you can be certain because it is unethical and dishonest. Your suggesting it only indicates to that you are ethically corrupt.

L: (Sigh) Yes, Steve, I can read - i know you aren't doing this. What I wondering is why *no one* is doing this. You answer appears to be "I don't know"?

Reply: You can't read very well either so I will go slow for you. Very carefully understand that there are hundreds of mediums/psychics and volunteers willingly involved in the such research, in the past and right now. I needn't bother some retired person who hardly gets out of her house to become a lab rat half a continent away. I dont think so. This is basic scumbag tactics and I would appreciate it if you would not fabricate phoney dialogue to go along with it also.

I brought up my experience, and I will go reaaallly slow here because I was asked a question about what got me interested and it was this person who represented the answer. I said it was anecdotal, I told you and anyone listening they could freekin take or leave it. But please don't harass this person any more. Thank you or we can start talking about your mom. Would you like that?

L: Yep. I await something interesting. Something beyond "statistical anomalies discovered through meta-analysis" would be good. Oh wait, you found a trance medium that is 100% accurate on almost 200 different items of information! How about we test that? Oh, sorry she's "too old" and "not interested". Okay.

reply: Its clear to me you havent lifted a book or read a journal paper on this subject so I feel you will be finding out about this when you are dead yourself and not before. You gotta get up off your arse and do some research. Again, please stop harassing this woman or I will be looking for information on Mrs. Loki to discuss as well. Was your mom named after an avenging angel also? Did she name you Loki for some special reason? Did she believe in angels Loki? You want more or will you stop harassing this woman?

L: Just because you want to accept anything and everything placed in front of you doesn't mean it does exist.

Reply: Unfortunately nothing is placed before me. One has to go out and find it. After you do, you can read, evauate and dscriminate. You, obviously, have done nothing in that respect or you would be discussing issues instead of acting like an armchair rhetorical skeptic.

This place and this discussion is a freeking waste of time. See you ina couple of days. I am taking a trip. Bye.

KelvinG

12th August 2003, 10:32 PM

Originally posted by SteveGrenard

This place and this discussion is a freeking waste of time. See you ina couple of days. I am taking a trip. Bye.

I don't think this is a very fair statement Steve. Loki made some good points. Just because he didn't nod obediently and agree with everything you said doesn't mean he isn't entitled to his opinion.

And do really expect anyone to believe that you know a medium who is 100% accurate on 200 pieces of information, yet she is suspiciously not available for testing.

Someone is naive and misinformed here Steve, but I don't think it's Loki.

SteveGrenard

12th August 2003, 10:56 PM

KG: I don't think this is a very fair statement Steve. Loki made some good points. Just because he didn't nod obediently and agree with everything you said doesn't mean he isn't entitled to his opinion.

Reply: Lets see why this is a waste of time:

1. Loki says we don't need the scientific method anymore, there is
non-scientific evidence to use. How convenient. But does he
offer it. no? We have ruled out anything from public performances as uncontrolled and unreliabe. Where is the smoking gun? If its Michael O'Neill, excuse me while I pick myself off the floor and try to stop laughing.

2. Loki says the established laws of physics preclude or somehow
prevents JE from being valid and causes him to be a fake.
I ask him to set forth all the established laws of physics and
what makes them established before I will go any further on
generalization. Its interesting how he nows segues back to
appealing to science again. He really does think he can have it
every way. Which is it going to be?

3. Loki invents fake dialogue between a private person (a trance
medium) and myself and ridicules her and me. I revealed my
experience with her in answer to a specific question (see
above). If nobody can answer a question here anymore
honestly without being ridiculed for it, I for one, feel not just
that this forum is a waste of time, it is not someplace where I
will, after this post, ever post again. There are other ports on
the internet where you will not be greeted by ridicule.
Where you can be skeptical and open minded and discuss real
facts instead of fantasies dreamed up by cynics to defend their
positions. I have also said, KG, in case you missed it about 5
times that my account was anecdotal and you can take it or
leave it. I really don't give a dam. That goes for you as well. Do what you want with that information.

KG: And do really expect anyone to believe that you know a medium who is 100% accurate on 200 pieces of information, yet she is suspiciously not available for testing.

Reply: You can think whatever you want. Other people here have been given information so they can see her for themselves.

KG: Someone is naive and misinformed here Steve, but I don't think it's Loki.

Reply: Think whatever you want KG. Its fine with me. I hope you and Loki are very happy together.

KelvinG

12th August 2003, 11:12 PM

Originally posted by SteveGrenard
Think whatever you want KG. Its fine with me. I hope you and Loki are very happy together.

Geez, it doesn't take you long to go from skeptical nice guy to angry believer.

But, just some friendly advice for the future. For someone who is supposedly so devoted to the scientific method and testing as proof of paranormal abilities, you may want to tone down the anecdotal evidence. Surely you realize you are opening yourself up to ridicule by making a seemingly outrageous claim and then not backing it up.
That's very unscientific if you ask me.

You remind me of a kid at school who tries to impress the other kids by bragging that his dad can lift a car above his head, but refuses to let anyone come over to his house and see the feat performed.
Are you trying to impress people Steve?
Do you think it's working?

You say you don't care about what people think and they can "take it or leave it" for all you care. Considering how angry you got in some of your previous posts, I think you do care what people think.
And you obviously don't take it very well when others disagree with you.

Anyways, good luck in the future Steve, and enjoy your trip.

Loki

12th August 2003, 11:22 PM

SteveGrenard,

Since you're away for a few days, there's no hurry to reply .. so I'll just touch on a copuple of points where we seem to be talking past each other.

(regarding dowsing)I cannot even begin to answer this question except to say I have no information, knowledge nor have I studied or been shown any evidence about dowsing so I cannot answer the question.
(sigh) Okay, well that about does it. You're either being deliberately obtuse, or you want me to believe that you don't know what dowsing is. What can I say??

...what other evidence is available that you cite? Please provide this.
...
Yes, what is the other evidence?
Already did, but you seem to want it repeated so at a minimum :

1. The complete lack of an "enabling mechanism" for ADC that is compatible with known physics.
2. Admitted fraud by other humans engaged in similar activity.

Look, my point is very simple ...

If you were a "skeptic" in the 15th Century you'd be entitled to say "the Earth is flat". That's a position that was supported by the available evidence :

1. The Sun clear moves across the sky;
2. If the earth was round, we'd start to walk "downhill" and eventually fall off if we travelled too far from home.

Since I can test #2 by simply walking for a few days in any direction, it's obviuous the Earth is flat.

If I was to meet a man claiming that the Earth was "cube shaped" (http://www.timecube.com) I'd be entitled to ask for evidence. If he couldn't provide any, I don't have to concede that he might be right simply because I haven't proven him wrong. If a few days later I meet a man who claims the Earth is "round", and he has evidence, I can look at his evidence and make up my own mind.

In relation to JE, this is very simple - he's making a specific claim, and (you admit this) providing absolutely no evidence whatsoever to support this claim. The fact that no one has proven his claim false doesn't automatically grant his claim "too close too call" status.

I think that I'm saying the Earth's round, and JE's trying to tell me it's a cube.

You think that I'm saying the Earth's flat, and JE's trying to tell me it's round.

Either way, the onus is on him to make his case!! If he does, I'll admit that my view of the Earth was mistaken, and I'll get used to living in a world shaped by JE. But until he backs up his claim, I don't have to give it equal weight!!! There is a body of evidence that says he's a fake.

You can't read very well either so I will go slow for you. Very carefully understand that there are hundreds of mediums/psychics and volunteers willingly involved in the such research, in the past and right now. I needn't bother some retired person who hardly gets out of her house to become a lab rat half a continent away. I dont think so
Perhaps we aren't speaking the same language, Steve. I'm not asking *you*, or your poor little old 100% accurate trance medium to get involved. And I'm not saying that no research is being done. I'm just asking why the believers themselves are not actively involved in research of their own. Claus is involved in the "Skeptic Report", and he (at least appears) to be interested in gathering and publishing data. Where's the believer equivalent? Where's the web sites where I can read about the "Hoboken Mediums Alliance", and the results of the trials they've conducted on their local medium? Every web site I visit seems to be full of quotes from books, and references to otehr web sites. Doesn't anyone bother to try and gather some data?

I was asking about "amateur" research. Both times you've answered" because *I* don't want to" - I get it, really I do. I though you might have some opinion on why this isn't widespread, given the enormous interest world wide in ADC - apparently not!

One last comment...
Is quantum theory turning physics on its rear end or not?

Quantum Mechanics makes my head hurt. Every time I think I've grasped some part of it, everything shifts a little to the left and I find myself examining a whole new set of concepts. But despite being such a wonderful field for woo-woo speculation, one thing seems clear (as best I can tell) - QM is not at odds with "classic" physics in the macro world. If you want to convince me that QM is somehow the mechanism for "psi" or "ADC", then produce the damn data, Steve. Until then, you're not "being a skeptic", you're just wishing and wanting. "Hey look everybody - there's an unexplored area of physics over there! Quick, lets theorise that psi probably exists in that bit we can't see!!!"

Loki

12th August 2003, 11:45 PM

SteveGrenard/KelvinG,

We'll probably cross post again (ah...3 sided conversations!), but I just wanted to comment on something Steve posted while I was preparing my last reply...

1. Loki says we don't need the scientific method anymore, there is
non-scientific evidence to use.
This comment tells me we aren't even close to understanding each other. I never said, or meant that. At the risk of labouring a point, I'd summarise what has happened like this :

Steve : JE must be either Debunked or validated using the scientific method before a reasonable person can have an opinon

Steve is claiming that JE *must* be debunked using the scientific method

Me : Not true - his claim is at odds with established evidence (namely physics and human nature)

Me saying that this is not true - not all claims must be debunked before an opinion can be formed

Steve : What! So know you say we don't have to use the scientific method? You say JE has to use to prove his claim, and you *don't* have to use it to prove yours!!

***BUZZ***

And that's where the error in understanding creeps in... The evidence I'm saying "works against" JE has been collected using the scientific method - it's the body of science itself that he's up against.

Gee, this is too hard, and probably not worth the effort. I only wanted to make a single point that seemed fairly clear cut - that not all claims are equal. To me, you debunk the data that someone puts up in support of their claim. For JE, there is no data. For the eleventy seventh time, it seems perfectly reasonable to me to reject JE until that data arrives - sitting on the fence on this issue is *not* the reasonable default position. I can see you disagree Steve, but I think that's just a weighting of the data, while you seem to think it's an abandoning of skeptical principles.

2. Loki says the established laws of physics preclude or somehow prevents JE from being valid and causes him to be a fake.
Exactly - the laws, arrived at via the scientific method, provide no know mechanism, or application of existing forces, that would produce ADC. Therefore, JE is either (a) a fake, (b) doing something that "beyond science (meaning science will never understand it)"; (c) exposing a flaw in an existing law; or (d) utilitising an entirely new law that is as yet undiscovered and undocumented.

The above choices *are* data derived logically from the laws of physics Steve, and they tell me that until such time as JE can provide supporting data, then (a) is far and a way the most likely alternative.

3. Loki invents fake dialogue between a private person (a trance medium) and myself and ridicules her and me. I revealed my experience with her in answer to a specific question (see
above).
Steve I apologise, since you feel this strongly about this - I though it was obvious that I was creating a parody to make a point. But one man's parody can be another's insult, so I'll just say that I didn't intend insult here - I thought it was humourous. Humour at your expense? Perhaps a little, but not intended to be taken so strongly. Sorry.

Now, I'll just post this, and read what was posted while I was writing!!!

(edited to add) : Hey whadda know - no cross posts!!

CFLarsen

13th August 2003, 12:14 AM

Originally posted by SteveGrenard
If nobody can answer a question here anymore honestly without being ridiculed for it, I for one, feel not just that this forum is a waste of time, it is not someplace where I will, after this post, ever post again.

OK, then. Ask Hal to have your account deleted. It's OK, I've saved all your posts anyway.

Originally posted by SteveGrenard
There are other ports on the internet where you will not be greeted by ridicule. Where you can be skeptical and open minded and discuss real facts instead of fantasies dreamed up by cynics to defend their positions.

Really? Pam's Board of Pam-Adoration? Or your own SurvivalScience? Oh, no, I forgot: You had to shut that down, because it turned out to be mainly Steve arguing with Steve.

It's been very interesting to know you. Most educational, in fact. You have done more to help skepticism on its way than few others, not by presenting a skeptical way of thinking, but by exposing just how silly, obtuse and ignorant a believer/wannabe-paranormal-researcher can be. You have made it abundantly clear that not even someone who dresses up in scientific mumbojumbo can stand the scrutiny of a few skeptics.

Maybe being a skeptic helps after all.

Buh-bye, Steve.

Lothian

13th August 2003, 01:07 AM

Originally posted by SteveGrenard
There are other ports on
the internet where you will not be greeted by ridicule.

http://www.woowoo.net
http://www.gullible.com and
http://intellectuallychallenged.com (http://www.intellectuallychallenged.com)

Kerberos

13th August 2003, 06:49 AM

Originally posted by renata

I can add all the names up- but if we want to know how many types of names, it will be more difficult, because there would be no exact overlap necessarily between 1900 and 1970, and I don't feel like calculating manually. But perhaps only the numbers of names in total for those years are relevant. In that case, I can do the math relatively easily- but I do not want to do it unless someone somewhere will use it for analysis! :)

And whereas we have this name data, it will be difficult to gather the untainted transcripts for analysis.

Well Claus has offered to find transcripts and I've agreed to count the names and letters.

I any case I think the US Census Bureau would be a better source, since we really don't know if the different letters occur as often in the 1000 most popular names as in the population total. The problem with that source of course is that it includes last names, and it’s my impression that JE mostly gets first names and in any case the entire family only a few different last names.

Jeff Corey

13th August 2003, 07:13 AM

Originally posted by SteveGrenard
LThis place and this discussion is a freeking waste of time. See you ina couple of days. I am taking a trip. Bye.
Don't take the brown acid!
too late.

Thanz

13th August 2003, 07:19 AM

Thanz

13th August 2003, 07:28 AM

Originally posted by BillHoyt

No, "suggestive" at least, not best. When scientists say "suggestive" they mean it lends support to an hypothesis. It is often used as an understatement.
No, I meant at best. In that, at best, it may lend support to the hypothesis that JE is cold reading. Given the state of the data used, it may just be useless. I don't think that it can be elevated beyond "suggestive", hence the use of the term "at best".

JE's initials and names refer to people both living and dead. Death records would not be more relevant.
Point taken.

How about you get more specific, Thanz? Why is it inadequate here?
How much more specific do I need to get? I don't think that the sample size is big enough. If we just use juninho's numbers for the sake of argument, some letters have a probability of about one in fifty. In 78 guesses, it is not anything meaningful if that letter does not appear. In 700 guess, that is much more relevant.

Let me ask it this way: If you picked 78 random people off the street, do you think that the name distribution will mirror that of the total population? Or do you think it is possible that it will be skewed? What if you took 1000 people at random, and compared their names? I think you would agree that the 1000 person sample is much more likely to mirror the name distribution of the population in general than the 78 person sample. I don't think that we could conclude much from the 78 person sample. It is just too small.

Lurker

13th August 2003, 07:32 AM

Originally posted by Thanz

I think you would agree that the 1000 person sample is much more likely to mirror the name distribution of the population in general than the 78 person sample. I don't think that we could conclude much from the 78 person sample. It is just too small.

Thanz is totally right here. I don't have the equation on me right now but there is one that governs confidence in sample size versus population size. Clearly 78 is far too small.

Lurker

BillHoyt

13th August 2003, 07:47 AM

Originally posted by Thanz
much more specific do I need to get? I don't think that the sample size is big enough. If we just use juninho's numbers for the sake of argument, some letters have a probability of about one in fifty. In 78 guesses, it is not anything meaningful if that letter does not appear. In 700 guess, that is much more relevant.

Let me ask it this way: If you picked 78 random people off the street, do you think that the name distribution will mirror that of the total population? Or do you think it is possible that it will be skewed? What if you took 1000 people at random, and compared their names? I think you would agree that the 1000 person sample is much more likely to mirror the name distribution of the population in general than the 78 person sample. I don't think that we could conclude much from the 78 person sample. It is just too small.
Thanz,

Try again. One cannot make the blanket statement that n=78 is inadequate unless one specifies the type of statistical test to be applied, the significance level to discriminate between accepting and rejecting the null hypothesis and the desired power of the test.

Cheers,

BillHoyt

13th August 2003, 07:49 AM

Originally posted by Lurker

Thanz is totally right here. I don't have the equation on me right now but there is one that governs confidence in sample size versus population size. Clearly 78 is far too small.

Lurker

Lurker,

What are your experimental design assumptions? What statistical test? What desired significance level? What power?

Cheers,

Thanz

13th August 2003, 07:56 AM

Originally posted by BillHoyt

Thanz,

Try again. One cannot make the blanket statement that n=78 is inadequate unless one specifies the type of statistical test to be applied, the significance level to discriminate between accepting and rejecting the null hypothesis and the desired power of the test.

Cheers,
Well, Mr. Hoyt, why don't you tell us for what set of parameters n=78 would be an adequate sample?

BillHoyt

13th August 2003, 08:02 AM

Originally posted by Thanz

Well, Mr. Hoyt, why don't you tell us for what set of parameters n=78 would be an adequate sample?
Thanz,

It seems you are willing to make the claim that n=78 is inadequate despite having no notion of the statistical test, the significance level or the desired power. Please prove me wrong by backing up your claim.

Cheers,

CFLarsen

13th August 2003, 08:03 AM

Originally posted by Thanz

Well, Mr. Hoyt, why don't you tell us for what set of parameters n=78 would be an adequate sample?

You can't say "No! Not good enough", and then not come up with something better.

If you think 78 is too small, then what is big enough? You obviously have worked this out statistically, so it should be easy for you to simply post your calculations.

BillHoyt

13th August 2003, 08:18 AM

Thanz,

Let's take this step-by-step. First question: what is your assumed statistical distribution model on which you base the claim n=78 is inadequate?

Cheers,

Thanz

13th August 2003, 08:41 AM

Mr. Hoyt and Mr. Larsen -

Before we go any further, do either of you think that the sample size of 78 IS adequate to give us actual meaningful, useful data in a comparison against letters of first names in the population at large? Is that your position - all we need is 78?

AlienX

13th August 2003, 09:13 AM

Originally posted by neofight

Don't mention it, AlienX. ;)

LOL, no, I'll spare you the rolleyes, AlienX. I realize that they are the only unedited readings that we have, but the reason that I don't believe it's fair to use those readings is because no matter what some might say, those of us who are familiar with JE's readings, from watching the show, and/or from going to seminars, know for a fact that these short little phone readings are not representative of a full JE reading. They're just not.

An average JE reading goes on for quite a while. These mini-readings are timed in mere seconds. They can't measure up to the usual length reading. It's impossible, and because of that fact, and because there is no immediate follow-up where the sitter can verify some of the things that JE said that flew by too fast for them to react to initially, these readings are less than ideal to work with. Far less than ideal......neo

edited to chance the word "edited" to "unedited"

I do have a couple of points RE: your reply.

JE does keep on going onto LKL and doing this so he obviously doesn't have a probelm with the conditions set. Doing this is not really publicity as the readings simply make you question his ability in the first place - so why do it?

If he is doing what he says he can then why is the information not accurate to begin with and apparently gets better the longer he spends with the sitter and if he can see them?. We have wild stabs in the dark that are blatantly incorrect then .. well you know the rest ;-)

OK to me any information given at the start of a reading MUST be as accurate as information given at the end of a reading - why would it be rubbish at the start given the claimed communication method?. To me the fact that a reading gets better over time points EVEN MORE towards cold reading. He is using information received to make deductions about the sitter - obviously the more information received the more accurate your guesses are going to be. That's what make Derren Brown (Not a psychic) so impressive - he gets awesome hits very quickly - I doub't he could do as well by phone though - you con't see the person - it's a nightmare scenario that noe for a cold reader.

Summary.

Given the claimed communication method why are short readings not accurate at all - really there should be no difference between accuracy between long and short readings - also a long reading there should be no difference between begining middle and end phases. Harder for JE to do this as he somethimes tries wild guesses to start with.

Better readings over a longer time frame reinforces cold reading as the mechanism. Short accurate readings with little to no feedback from the reader would reinforce "real" mediumship but with JE we are not seeing this pattern.

TA

AlienX

BillHoyt

13th August 2003, 09:21 AM

Originally posted by Thanz
Mr. Hoyt and Mr. Larsen -

Before we go any further, do either of you think that the sample size of 78 IS adequate to give us actual meaningful, useful data in a comparison against letters of first names in the population at large? Is that your position - all we need is 78?
Thanz,

You have claimed it is not. Please answer the questions put to you:

o what distribution model do you assume?
o what significance level do you want to use to discriminate between accepting and rejecting the null hypothesis?
o what power do you want to achieve with the test?

Cheers,

CFLarsen

13th August 2003, 09:26 AM

Originally posted by Thanz
Mr. Hoyt and Mr. Larsen -

Before we go any further, do either of you think that the sample size of 78 IS adequate to give us actual meaningful, useful data in a comparison against letters of first names in the population at large? Is that your position - all we need is 78?

Mr. Thanz,

No, not "before we go any further". You claimed that 78 was inadequate. Now, tell us what you consider adequate, and why.

Please refer to the appropriate statistical methods. It must be easy, since you were so quick to complain. No more stalling, no more excuses.

Thanz

13th August 2003, 09:30 AM

Mr. Hoyt -

I have no desire to go through a statistical analysis of something that I feel is obvious unless there is an actual dispute. If you do not think that 78 is adequate, then we have no dispute.

So, before we go through any detailed analysis, is it your position that 78 is adequate to give us actual meaningful, useful data in a comparison against letters of first names in the population at large? Is that your position - all we need is 78?

I have no desire to argue for the sake of arguing.

Lurker

13th August 2003, 09:36 AM

Originally posted by BillHoyt

Lurker,

What are your experimental design assumptions? What statistical test? What desired significance level? What power?

Cheers,

Experimental Design: Let us assume you can randomly choose the samples from the total population. By doing this, you are assuming a best possible scenario. Anything else would require a LARGER sample size to make up for the deficiency.

Statistical Test: http://www.surveysystem.com/sscalc.htm

Input 95% confidence in your data. Then a confidence interval of 5, which is VERY high for the multivariate statistics you are using here. With so many names, the +-5% for each name would be exceedingly high. Next, enter the population fo 280million. Have it calculate and voila! Out pops the number 384 for a sample size. Far higher than 78, right?

I agree it is not exact and leaves a bit to be desired but it is a decent approximation, no?

Lurker

CFLarsen

13th August 2003, 09:40 AM

Originally posted by Thanz
Mr. Hoyt -

I have no desire to go through a statistical analysis of something that I feel is obvious unless there is an actual dispute. If you do not think that 78 is adequate, then we have no dispute.

So, before we go through any detailed analysis, is it your position that 78 is adequate to give us actual meaningful, useful data in a comparison against letters of first names in the population at large? Is that your position - all we need is 78?

I have no desire to argue for the sake of arguing.
It is most definitely not for the sake of arguing. You complained about 78 not being adequate. Before we go on with this analysis, we should most definitely hear from you, who are obviously an expert in these matters.

Please stop stalling. Stop evading. Stop changing the subject.

Or admit that you have no idea what the heck you are talking about.

Thanz

13th August 2003, 09:43 AM

I agree with Lurker.

Edited to add:

Oh, and Claus, i think it is indeed arguing for the sake of arguing. Otherwise, why the reluctance on the part of both Mr. Hoyt and yourself to commit to 78 as an adequate number?

CFLarsen

13th August 2003, 10:04 AM

Originally posted by Thanz
I agree with Lurker.

Edited to add:

Oh, and Claus, i think it is indeed arguing for the sake of arguing. Otherwise, why the reluctance on the part of both Mr. Hoyt and yourself to commit to 78 as an adequate number?
Thanz,

I think it is obvious by now that you had no idea what you were talking about. You have stalled, evaded, obfuscated this issue. Tried to put the onus on others, shifting the blame. Just about all those techniques we see from woowoos.

Now, you lean on Lurker, instead of coming up with something yourself. You agree, but you don't explain why you agree. You simply don't have a clue what Lurker is talking about, do you?

You do that all the time: You never do anything yourself, all you do is complain about others.

Now, it's all BillHoyt and my fault: We are - supposedly - the ones who are "reluctant". No, Thanz: You are the one who is "reluctant".

Some might even say frantically on the run.

Thanz

13th August 2003, 10:14 AM

Originally posted by CFLarsen

Thanz,

I think it is obvious by now that you had no idea what you were talking about. You have stalled, evaded, obfuscated this issue. Tried to put the onus on others, shifting the blame. Just about all those techniques we see from woowoos.

Now, you lean on Lurker, instead of coming up with something yourself. You agree, but you don't explain why you agree. You simply don't have a clue what Lurker is talking about, do you?

You do that all the time: You never do anything yourself, all you do is complain about others.

Now, it's all BillHoyt and my fault: We are - supposedly - the ones who are "reluctant". No, Thanz: You are the one who is "reluctant".

Some might even say frantically on the run.
Oh please, give it a rest Claus.

I stated something that I think is glaringly obvious - if you want to examine JE guesses of letters of the alphabet in comparison to the general population, you are going to need more than 78 guesses to work with. Lurker has kindly confirmed this.

As for whether or not I do anything myself, I did find a source of names for this very analysis.

Both you and Mr. Hoyt have jumped on me to prove my claim - which, as I said, I view as glaringly obvious - while carefully using language that doesn't actually dispute my claim. When asked to actually dispute my claim, you decline.

If you can come up with a reason to prove my claim when you haven't even disputed it (and in fact have refused to dispute it) I'd love to hear it.

Of course, I just think that you and Mr. Hoyt refused to dispute my claim because even you know that 78 is not adequate - and that its inadequacy is glaringly obvious.

renata

13th August 2003, 10:24 AM

Originally posted by Thanz

Maybe I am just being dense, but I don't know what you mean by "types of names". I am not sure what your difficulty is. Can you explain it so that even I can understand?

No, it was late and it is hard for me to write clearly at night (sounds of- when do you ever write clearly cat calls)

So, in the 1900 top 1000 name example, we counted
97 A names, for total of 31,584 names
51 B names for 7,734 names
71 C names, for 35,915 names

If we were to combine all years, the number of name types would vary from decade to decade as names get disused or appear. So in 1970 we coul have- for example (made up numbers)
88 A names for total of 56,000 names
63 B names for a total of 9,076 names
88 C names for a total of 56,823 names

Whereas we can add the name totals from decade to decade, and that will give us a good number, we will not know how many name A names, B names etc exist, because they will not overlap exactly decade to decade

Am I making ANY sense here? :)

Originally posted by Kerberos
Well Claus has offered to find transcripts and I've agreed to count the names and letters.

I any case I think the US Census Bureau would be a better source, since we really don't know if the different letters occur as often in the 1000 most popular names as in the population total. The problem with that source of course is that it includes last names, and it’s my impression that JE mostly gets first names and in any case the entire family only a few different last names.

Kerberos, if you wish to do the count, help yourself! :)

I think both sources are useful. The source Thanz provided shows the amount of names given to children for a given decade. You could use it to calculate all names given to children born in the US for the whole century. That should cover dead and living, parents and grandparents. I do not think JE connects with anyone who was born prior to 1900 (besides, those people would have been reincarnated according to him but it is another story altogether)

The 1990 census will only show names of people living in US that year. However, it will also have names of immigrants, which the baby index will not.

I will stay out of this unless you guys want me to assist, then :)

CFLarsen

13th August 2003, 10:24 AM

Thanz,

You wilfully misunderstand. Yes, wilfully, because you have been avoiding the issue like the plague.

We are not saying that 78 is perfect. What you need to explain is:

What is your assumed statistical distribution model on which you base the claim n=78 is inadequate?
What distribution model do you assume?
What significance level do you want to use to discriminate between accepting and rejecting the null hypothesis?
What power do you want to achieve with the test?

No, it's not about me, BillHoyt or anybody else but you. You say 78 is not adequate.

Why not?

In statistical terms, please. No leaning on the works of others.

And no, I will not "give it a rest". I just want to hear why you think it is "obvious" that 78 is inadequate.

Thanz

13th August 2003, 10:42 AM

Claus -

I have already explained why I thin 78 is inadequate. I note that in your latest response, you have stated that 78 is "not perfect". That was not my question. My question was whether it was adequate for any meaningful purpose.

If you do not think it is adequate for any meaningful purpose, I see no point in continuing to argue with you. If you do, I want you to actually say it.

From my perspective, it is obvious to all that 78 is not enough. Including you. But you, because you do not like me, have decided that this would be a good issue to ride me for some reason. Fine, go ahead. But at least have the guts to tell me I'm wrong first. Which both you and Hoyt have declined to do.

CFLarsen

13th August 2003, 10:48 AM

Thanz,

Liking you or not has nothing to do with it. And nobody is saying you are wrong.

You made a claim. You back it up. Those are the rules of this board.

What is your assumed statistical distribution model on which you base the claim n=78 is inadequate?
What distribution model do you assume?
What significance level do you want to use to discriminate between accepting and rejecting the null hypothesis?
What power do you want to achieve with the test?

These questions will reappear until you either:

1) Answer them.

or

2) Admit that you don't want to answer them.

Leroy

13th August 2003, 10:52 AM

by cflarsen - Perhaps, in time, you will actually contribute with something valuable to the discussions here

This, spoken by the man who spends his time writing, and re-writing, and writing again, that one sentence "YOU ARE A LIAR!" without ever proving a darned thing. :dl:

Thanz

13th August 2003, 10:58 AM

Originally posted by CFLarsen
Thanz,

Liking you or not has nothing to do with it. And nobody is saying you are wrong.
Have YOU done the analysis that you are asking me to perform? No? Then why are you asking me to do it? You do not dispute my conclusion. You have done the same thing that I have done.

These questions will reappear until you either:

1) Answer them.

or

2) Admit that you don't want to answer them.
I will not be answering these questions. I don't see the point. You say that "Nobody is saying that I am wrong". Will I have to prove that the sky is blue next?

At what point does it become ridiculous to even ask me these questions? If I said that 50 guesses were inadequate, would you ask the same thing?

voidx

13th August 2003, 11:00 AM

Posted by neofight:
Exactly, renata, and I'm so pleased to see that you understand this concept. You now understand precisely how John Hockenberry picked and chose exactly what he would air on his "Dateline" show so that he would make John Edward appear to have purposely cheated on national television. Thanks! You're a peach! ......neo
Yes, you have a contrary example in your opinion of Hockenberry doing the same thing with editing, this is now established. Now, answer Renata's question and quit deflecting.

Posted by neofight:
Hi, Leroy! I just would like to clarify something here. He does not generalize here and just throw in "suicide", but only a drug overdose, accidental or otherwise, that would indicate a lot of "toxins" in the blood.
And toxins are not disorders. We all understand how he is able to get away with it. Now lets all acknowledge that it is, intentional or not(I believe intentional), that it is an overly vague category.

Posted by neofight:
An average JE reading goes on for quite a while. These mini-readings are timed in mere seconds. They can't measure up to the usual length reading. It's impossible, and because of that fact, and because there is no immediate follow-up where the sitter can verify some of the things that JE said that flew by too fast for them to react to initially, these readings are less than ideal to work with. Far less than ideal......neo
To ask again, why doesn't JE object then? You can hypothesize all sorts of things from doing it for exposure, or publishers making him actively do it to promote new books, but at the end of the day, it makes him look bad. You seem to be admitting in a roundabout way that JE's LKL readings are in fact not very good, will you admit that? If they make him look bad, it would be in his interest not to do them, or to change the format so that he could do them better, yet he doesn't. And as for the contention that a 1 or 2 minute reading couldn't provide good results I present you with this post by instig8r quoting JE:

Posted by Instig8r::
Here is an excerpt from "One Last Time", written by John Edward (and posted by Neo over at TVTalk last year):

"All of us--we in physical bodies and those in the spirit world--are made up of energy expressed as atoms and molecules spinning and vibrating at certain speeds. The energy of spirits vibrates at a very high rate, while ours goes much slower because we are in physical bodies. How we bridge the gap dictates how well communications traverse these two dimensions. That's the job of a medium.

For spirits to come through, they must slow their vibrational rate of energy. Think of the blades on a helicopter. You can't see that there are four of them because they are spinning too fast. That's how I view the energy of the spirits. What happens during a reading is that as the spirits slow down their energies, I speed mine up. Communication is what happens in that space in between. But because there is that space, that gap, communication is never easy and rarely clear. There is also the fact that spirits no longer have physical bodies to facilitate communication. They have no tongues or vocal cords to pronounce words. Instead, through their energies, they place thoughts and sights and sounds in my mind. I am their mouthpiece. Though I expected to hear a great voice from beyond when I first started this work, I soon realized that it is my own voice I hear--but their thoughts and feelings.

Because both sides must expend so much energy to make this happen, the communication is very difficult and can't be sustained for more than a few minutes. To switch metaphors, it's as if you have to go to the bottom of a twelve-foot-deep swimming pool to meet your loved one. You can do it, but it takes a lot of energy to get there and after a few seconds, you have to float back to the top for air."

From the horses mouth. So if it can only be sustained for a few minutes how does he manage these long extended readings on CO? Like I've always stated, he's not even consistent in his own method, or how he describes it.

Loki, I'm certain if he was asked the question, he would have to agree that having the people he is reading right there in front of him has to be less confusing to him than having ten people holding on phone lines.

I mean, it's common sense. At least when they are in front of him physically, he feels a "pull" as to whom he is supposed to be with that leads hiim to the right area. I doubt that happens on the phone......neo
I will try and find it, but it was another quote from JE, he's said that communicating with spirits and doing a reading isn't affected by distance, or by being on radio, or the telephone or TV. So the first paragraph seems invalid by JE's own description of what he does. As for the second part, you see it as a "pull" to make a proper connection, something paranormal. We would see it as reading body language, to have people physically in front of him is to read reactions to info being tossed out, of course its easier, for a cold-reader. A cold-reader would do better when sitting right in front of the person because they'd be able to judge body language. JE also seems to do better when sitting physically in front of people. Coincendence maybe, but I'd lean towards not.

Posted by Steve Grenard:
Whoever, said JE has control over the calls is mistaken. King has a switch. When the call goes over, even if the caller or JE is not finished., he (King himself) cuts them off. Next caller. Cut. Next caller. Cut. Next caller. Cut.
Obviously Larry has final control over the calls, and no one that I remember stated that JE has control over them. I myself mentioned that its not supported by the transcripts that JE is cut off repeatedly while trying to read. You say he cuts them off, can you show transcripts that support this?

CFLarsen

13th August 2003, 11:11 AM

Originally posted by Thanz
Have YOU done the analysis that you are asking me to perform? No? Then why are you asking me to do it? You do not dispute my conclusion. You have done the same thing that I have done.

Funny, coming from you. You, who have made it your trademark to demand that others do your homework.

Originally posted by Thanz
I will not be answering these questions. I don't see the point. You say that "Nobody is saying that I am wrong". Will I have to prove that the sky is blue next?

No, all you have to do is answer the questions regarding why you think 78 is inadequate. Merely saying "I agree with Lurker" is not enough.

Originally posted by Thanz
At what point does it become ridiculous to even ask me these questions? If I said that 50 guesses were inadequate, would you ask the same thing?

Asking people to back up their claims can never be considered ridiculous.

Do you refuse to answer the questions then?

Lurker

13th August 2003, 11:12 AM

Where did the 78 number come from anyway?

Lurker

Lurker

13th August 2003, 11:16 AM

Never mind, I went back and looked again where teh 78 came from.

Lurker

BillHoyt

13th August 2003, 11:22 AM

Originally posted by Thanz
I will not be answering these questions. I don't see the point. You say that "Nobody is saying that I am wrong". Will I have to prove that the sky is blue next?

At what point does it become ridiculous to even ask me these questions? If I said that 50 guesses were inadequate, would you ask the same thing?

Here is the point, Thanz: you have no idea what you are talking about. That is abundantly clear. You have specified nothing about the experimental design and yet state, uncategorically, that 78 is obviously inadequate.

Thanz

13th August 2003, 11:26 AM

Originally posted by CFLarsen
Funny, coming from you. You, who have made it your trademark to demand that others do your homework.
Simply pointing out the hypocrisy, Claus. You haven't done the analysis that you are asking me to, yet you come to the same conclusion. Apparently you do not expect from yourself what you demand of others.

No, all you have to do is answer the questions regarding why you think 78 is inadequate. Merely saying "I agree with Lurker" is not enough.
I have already explained why I think 78 is not enough.

Asking people to back up their claims can never be considered ridiculous.
I disagree, when the claim is something that should be obvious.

Let's say that we only had 10 of JE's guesses, and I said that 10 guesses is woefully inadequate, would you be asking me these questions? Wouldn't it be ridiculous to ask me these questions?

Do you refuse to answer the questions then?
Hmmm.... In my last post I stated "I will not be answering these questions"...... Let's see if you can figure it out.

BillHoyt

13th August 2003, 11:29 AM

Originally posted by Thanz
From my perspective, it is obvious to all that 78 is not enough. Including you. But you, because you do not like me, have decided that this would be a good issue to ride me for some reason. Fine, go ahead. But at least have the guts to tell me I'm wrong first. Which both you and Hoyt have declined to do.
Thanz,

Wow. You don't even understand the relevancy of the questions, do you? You try to cover this with an appeal to popularity and then a subject/motive shift.

Here is the answer, Thanz: until you state your statistical test assumptions, nobody can know whether 78 is adequate or not. Set the significance level lower and 78 can be adequate. Set it higher and it can be inadequate. Set the required power lower and it can be adequate. Set it higher and it can be inadequate. You don't even understand the basic statistical questions or their relevancy, but have the audacity to make statistical claims!

Wow.

Thanz

13th August 2003, 11:31 AM

Originally posted by BillHoyt
Here is the point, Thanz: you have no idea what you are talking about. That is abundantly clear. You have specified nothing about the experimental design and yet state, uncategorically, that 78 is obviously inadequate.
Here is the point, Mr. Hoyt: it IS obviously inadequate. No one has even tried to say that it is adequate. Not you, not Claus, not anyone. I am forced to ask "why?" Why is it, if I know nothing about what I am talking about, that no one even disputes what I am saying? If I am so glaringly deficient, why can't you even make the assertion that I am wrong?

And if my assertion is not wrong, why are you hounding me?

BillHoyt

13th August 2003, 11:34 AM

Originally posted by Thanz

Here is the point, Mr. Hoyt: it IS obviously inadequate. No one has even tried to say that it is adequate. Not you, not Claus, not anyone. I am forced to ask "why?" Why is it, if I know nothing about what I am talking about, that no one even disputes what I am saying? If I am so glaringly deficient, why can't you even make the assertion that I am wrong?

And if my assertion is not wrong, why are you hounding me?

Read the second post and try again.

Lurker

13th August 2003, 11:39 AM

Actually, a good question might be why is the 78 number adequate? Clearly someone floated the number originally. That person should then defend its accuracy, no?

If I have a beef with it, is it MY duty to explain why? If the number 78 was originally floated with no support whatsoever then it is their duty to defend it.

Now neither Cantata nor Hoyt floated the number originally so are not responsible for defending it. Yet why did they not question the 78 number in the first place if they are so interested in evidence?

I am not saying that my method is definitive. It is a gross approximation but it is better than any support for the number 78 that I have seen yet.

As to Thanz, he is clearly saying that he felt it was obvious that 78 was too small. He makes no claims to a knowledge of statistics but appeals to common sense which nobody has called into question. Fine by me. If anyone feels otherwise, present your evidence.

Lurker

BillHoyt

13th August 2003, 11:43 AM

Originally posted by Thanz
And if my assertion is not wrong, why are you hounding me?

Since you are still not getting it, let me try to be crystal clear. Your assertion IS wrong.

Now will you be so kind as to try to qualify your assertion, in proper statistical terms to attempt to make it right?

Do you get it now?

BillHoyt

13th August 2003, 11:46 AM

Originally posted by Lurker
Actually, a good question might be why is the 78 number adequate? Clearly someone floated the number originally. That person should then defend its accuracy, no?

If I have a beef with it, is it MY duty to explain why? If the number 78 was originally floated with no support whatsoever then it is their duty to defend it.

Now neither Cantata nor Hoyt floated the number originally so are not responsible for defending it. Yet why did they not question the 78 number in the first place if they are so interested in evidence?

I am not saying that my method is definitive. It is a gross approximation but it is better than any support for the number 78 that I have seen yet.

As to Thanz, he is clearly saying that he felt it was obvious that 78 was too small. He makes no claims to a knowledge of statistics but appeals to common sense which nobody has called into question. Fine by me. If anyone feels otherwise, present your evidence.

Lurker

78 was not "floated", lurker. It was the number available from the available, unedited transcripts. Thanz objected to it. You now echo that on "common sense" grounds. Give us all a break.

CFLarsen

13th August 2003, 11:47 AM

Originally posted by Thanz
Simply pointing out the hypocrisy, Claus. You haven't done the analysis that you are asking me to, yet you come to the same conclusion. Apparently you do not expect from yourself what you demand of others.

No analysis is needed, just the statistical explanation that caused you to state it.

Originally posted by Thanz
I have already explained why I think 78 is not enough.

No, you have merely said you agreed with Lurker. No explanation from you whatsoever.

Originally posted by Thanz
I disagree, when the claim is something that should be obvious.

But why is it obvious? That's what you need to explain. But won't.

Originally posted by Thanz
Let's say that we only had 10 of JE's guesses, and I said that 10 guesses is woefully inadequate, would you be asking me these questions? Wouldn't it be ridiculous to ask me these questions?

No, because you claimed there was a statistical reason.

Originally posted by Thanz
Hmmm.... In my last post I stated "I will not be answering these questions"...... Let's see if you can figure it out.

Very well: You have refused to answer the questions. You made a claim, and were unable to back it up.

Lurker

13th August 2003, 11:54 AM

Originally posted by BillHoyt

78 was not "floated", lurker. It was the number available from the available, unedited transcripts. Thanz objected to it. You now echo that on "common sense" grounds. Give us all a break.

Oh, I understand, but implicit in the 78 was whether the sample size was large enough or not to represent the population. Thanz posited that it is not.

Since nobody seems to be defending the number 78, why is everyone arguing about it? It really is a minor point.

Further, when I saw the number I thought it too small as well. Luckily for me I DO have some background in statistics and probability. So I used common sense and when you asked me, I brought in some statistics to refute the number.

So now, we have my weak refutation versus no support. At least I have some evidence showing 78 is too small.

If you like, feel free to provide evidence to the contrary. Otherwise, you are starting to sound a bit like a woo-woo in defending something with no support.

Lurker

Of course, I could be wrong...

CFLarsen

13th August 2003, 11:58 AM

Originally posted by Lurker
As to Thanz, he is clearly saying that he felt it was obvious that 78 was too small. He makes no claims to a knowledge of statistics but appeals to common sense which nobody has called into question.

Actually, he did claim knowledge of statistics:

"Next, 78 guesses is a woefully small sample size. I don't know if we can really glean anything significant from such a small sample."

"If you picked 78 random people off the street, do you think that the name distribution will mirror that of the total population?"

"I think you would agree that the 1000 person sample is much more likely to mirror the name distribution of the population in general than the 78 person sample. I don't think that we could conclude much from the 78 person sample. It is just too small."

Thanz is clearly referring to statistics here. But, as we have seen, he cannot explain what he meant. We can only conclude that he had no idea what he was talking about.

Leroy

13th August 2003, 12:01 PM

Posted by Neo - Hi, Leroy! I just would like to clarify something here. He does not generalize here and just throw in "suicide", but only a drug overdose, accidental or otherwise, that would indicate a lot of "toxins" in the blood.

I understand what you are saying, but I think that he can use the term "blood" to fit almost anything if he wants.

J- I am seeing something to do with the blood
Sitter - my father was in a car accident
J- did he bleed?

J - I am seeing blood, that means something affected the blood
Sitter - my son died of a drug overdose
J - okay, toxins in the blood

J - something affected his blood?
Sitter - He died when they were doing a heart transplant
J - Ah, okay, he lost a lot of blood then

J - something affected his blood?
Sitter - He was hit by a car
J- did they give him blood while he was in the hospital
Sitter - no
J- did they take blood while he was in the hospital

He didn't actually make these statements, these are my examples - but if you think about it, one could make that word fit so many different causes of death.

I am curious as to other things he sees as references to how someone died.

Posted by Bill - Stop wasting our time and JREF bandwith!

:roll: You waste entire threads Bill - threads dedicated to finding and pointing out flaws in statements made by posters you don't happen to like, so who are you to tell Neofight to stop wasting JREF badnwith.

So, if we can't trust JE's readings on LKL, or his readings on CO, then why is there even discussion about him.

If someone is on national television making millions of dollars a year claiming to speak with the dead, that makes a good discussion.

Discussing it can help others who are interested in the show, see viewpoints from the believer and the skeptic. It is necessary to discuss it for this reason. What if we didn't discuss him. What if the only sites dedicated to discussing John Edward were sights that promote him? I think it is good to have this topic on a skeptics board.

Good post Kevin

If commonsense was applicable, then surely JE would be able to get letters (even words) from the spirits - they are, after all, the simplest and most common 'symbols' within the 'frame of referecen' of JE, the spirit, and the sitter

Yeah, what he said!

BillHoyt

13th August 2003, 12:01 PM

Originally posted by Lurker

Oh, I understand, but implicit in the 78 was whether the sample size was large enough or not to represent the population. Thanz posited that it is not.

Since nobody seems to be defending the number 78, why is everyone arguing about it? It really is a minor point.

Further, when I saw the number I thought it too small as well. Luckily for me I DO have some background in statistics and probability. So I used common sense and when you asked me, I brought in some statistics to refute the number.

So now, we have my weak refutation versus no support. At least I have some evidence showing 78 is too small.

If you like, feel free to provide evidence to the contrary. Otherwise, you are starting to sound a bit like a woo-woo in defending something with no support.

Lurker

Of course, I could be wrong...
Lurker,

To claim 78 is inadequate is impossible without qualifying the claim by answering the questions I posed. You claim you brought in "some statistics" to refute the number. No you did not. You merely alluded to confidence intervals and population size. You did not answer any of the essential questions. Apparently, you don't understand why they are essential despite your claimed background in statistics.

Cheers,

BillHoyt

13th August 2003, 12:06 PM

A funny thought just occurred to me. The nature of the Laudies is such that they credit multiple posters for their work. Perhaps I need to also open up the Totles to group efforts. Lurker and Thanz are working their way into the first group nomination...

Cheers,

Instig8R

13th August 2003, 12:14 PM

Originally posted by neofight

LOL, 'g8R! Are you serious? Regardless of what you may have said afterwards, that does not alter the fact that you told
Kerberos quite directly NOT to listen to me about the editing!!! :rolleyes:

As I said, it was not your disagreeing with me that I took exception to, it was your taking it upon yourself to instruct Kerberos to disregard my post, as though your opinion was worth considering, but mine was to be dismissed. That's all.

Neo-- In the past, you have complained about other posters being too LITERAL in their interpretations. Don't you think you are being unnecessarily literal in your interpretation of my comment?

If you are now the self-appointed "Hall Monitor" of JREF, please tell me how I could have phrased that post to please you. Shall I adopt the style that you often use when disputing a point that I make? For example, over at TVTalk, you not only took it upon yourself to tell others to disregard my posts,but you also called me dishonest and unfair.

I was not attempting to deflect attention from anything, since I've often stated that I disagree with the significance you attach to those types of descrepencies that involve interpretation of symbols. We've gone over that many times. I know we disagree, and I accept that, no problem.

And I did not make up accusations of hatreds and biases about you. If you think *hate* is too strong a word to use, and perhaps you're right, it might be, then substitute *strong dislike* and/or *contempt*, and perhaps that would be more accurate.

Neo, there is no reason for you to attach any emotional state or try to presume what another poster is feeling. It is simply not germane to the issue. You should deal only with the content of the post.

If you do not have a valid argument against what someone says, there is nothing to be gained by trying to discredit someone's opinion by accusing them of having sinister motives and character flaws.

With regard to the bias that I think you demonstrate against JE, there we will have to agree to disagree I think, because imo, your bias is not exactly understated. Come on now! I admit to my bias, 'g8R! Please don't deny yours. ;) .....neo

Uh, Neo, it is not necessary for you to continuously bring up the topic of bias whenever there is a debate. In the scientific community, there is an automatic presumption of bias. That is why research studies are often "double-blind"-- to rule out bias.

However, we were not talking about bias. You embarked upon full-scale ad hom against me. Without any basis in fact, you malaigned me by claiming that I have a malicious, obsessive hatred of JE. You stated that I have used nasty, hateful rhetoric when discussing JE -- and that other people have commented on my malicious language.

Again, this is NOT about bias... This is about your gratuitous, unsubstantiated accusations of malice.

I'm still waiting to see the evidence...

BillHoyt

13th August 2003, 12:19 PM

Originally posted by Leroy
You waste entire threads Bill - threads dedicated to finding and pointing out flaws in statements made by posters you don't happen to like, so who are you to tell Neofight to stop wasting JREF badnwith.
Leroy,

JREF is an educational foundation. Its mission is to spread critical thinking about the paranormal and pseudoscience. The foundation of critical thinking is two-fold: getting the facts straight and thinking logically about them. Someday you'll get it. But not today.

voidx

13th August 2003, 12:29 PM

Originally posted by BillHoyt

Thanz,

Wow. You don't even understand the relevancy of the questions, do you? You try to cover this with an appeal to popularity and then a subject/motive shift.

Here is the answer, Thanz: until you state your statistical test assumptions, nobody can know whether 78 is adequate or not. Set the significance level lower and 78 can be adequate. Set it higher and it can be inadequate. Set the required power lower and it can be adequate. Set it higher and it can be inadequate. You don't even understand the basic statistical questions or their relevancy, but have the audacity to make statistical claims!

Wow.
I don't want to get involved period but what your saying, that perhaps lurker and thanz are missing is that whatever weight, or significance you give to each reading, directly effects how adequate or inadequate the total of 78 readings is correct? If so then explain to us on what basis we set this significance. I'm not saying you're wrong, but Lurker and Thanz aren't getting what you're trying to tell them, so why not explain it out in detail for us, so that it eliminates any confusion. Otherwise we're going to run around in circles all day...again.

Leroy

13th August 2003, 12:31 PM

Posted by Thanz - Well, Mr. Hoyt, why don't you tell us for what set of parameters n=78 would be an adequate sample?

Posted by BillHoyt - It seems you are willing to make the claim that n=78 is inadequate despite having no notion of the statistical test, the significance level or the desired power. Please prove me wrong by backing up your claim.

hmm, I translate this to mean that Bill cannot answer your question.

BillHoyt

13th August 2003, 12:36 PM

Originally posted by voidx

I don't want to get involved period but what your saying, that perhaps lurker and thanz are missing is that whatever weight, or significance you give to each reading, directly effects how adequate or inadequate the total of 78 readings is correct? If so then explain to us on what basis we set this significance. I'm not saying you're wrong, but Lurker and Thanz aren't getting what you're trying to tell them, so why not explain it out in detail for us, so that it eliminates any confusion. Otherwise we're going to run around in circles all day...again.

voidx,

That is part of it, but not the whole story. Thanz is the one who said 78 was inadequate. He could only have made that claim by:

o having a distribution model, significance level and power in mind, or by
o pulling it out of thin air, with no foundation whatsoever.

I'll play along though. Let us say the significance level we desire is .05. Now, Thanz, tell us the rest.

Cheers,

CFLarsen

13th August 2003, 12:37 PM

Originally posted by Leroy
hmm, I translate this to mean that Bill cannot answer your question.

I translate this to mean that Thanz tries to get other people to answer his claim for him.

I also translate your post to mean that you ignore this.

Do you bring anything to the table yourself, Leroy?

BillHoyt

13th August 2003, 12:38 PM

Originally posted by Leroy
hmm, I translate this to mean that Bill cannot answer your question.

I answered the question, Leroy. You missed it because you are out of your league here. But that's been obvious for quite some time.

Set yourself on an easter egg hunt. You may find the answer.

voidx

13th August 2003, 12:47 PM

Originally posted by BillHoyt

voidx,

That is part of it, but not the whole story. Thanz is the one who said 78 was inadequate. He could only have made that claim by:

o having a distribution model, significance level and power in mind, or by
o pulling it out of thin air, with no foundation whatsoever.

I'll play along though. Let us say the significance level we desire is .05. Now, Thanz, tell us the rest.

Cheers,
I'm not interested in proving Thanz or anyone else wrong, and I could care less who or why someone said 78 was inadequate. You realize that if you state what level of significance we should be using for a sample size of 78 in this case, and why we should be using that particular significance level, then we can decide that 78 is indeed adequate and move on, or we can find another number. In your opinion Thanz and Lurker don't understand what you're getting at and couldn't explain why 78 is adequate, so rather than go out of you're way to stump them, why not explain it to us so we can move on. Its been years since I took any statistics so I can't be of any help, but between you, CFL, Lurker and Thanz you should be able to come up with something that makes sense, just quit trying to score points for like 10 minutes is all I'm asking...of everyone.

BillHoyt

13th August 2003, 12:59 PM

Originally posted by voidx

I'm not interested in proving Thanz or anyone else wrong, and I could care less who or why someone said 78 was inadequate. You realize that if you state what level of significance we should be using for a sample size of 78 in this case, and why we should be using that particular significance level, then we can decide that 78 is indeed adequate and move on, or we can find another number. In your opinion Thanz and Lurker don't understand what you're getting at and couldn't explain why 78 is adequate, so rather than go out of you're way to stump them, why not explain it to us so we can move on. Its been years since I took any statistics so I can't be of any help, but between you, CFL, Lurker and Thanz you should be able to come up with something that makes sense, just quit trying to score points for like 10 minutes is all I'm asking...of everyone.
voidx,

And I'm not interested in doing Thanz' work for him! He made the claim, and I have called on him to support it. He has now hemmed and hawed and, ultimately, refused to answer the questions. I have also answered the question on whether or not 78 is adequate. And part of the point, here, voidx, is the very antithesis of your very last line. Thanz and Lurker want us to believe they can come up with something. So far, that something is hot air, a drive-by assertion that 78 is "obviously inadequate".

Cheers,

Lurker

13th August 2003, 01:35 PM

Originally posted by BillHoyt

Lurker,

To claim 78 is inadequate is impossible without qualifying the claim by answering the questions I posed. You claim you brought in "some statistics" to refute the number. No you did not. You merely alluded to confidence intervals and population size. You did not answer any of the essential questions. Apparently, you don't understand why they are essential despite your claimed background in statistics.

Cheers,

Essential question as defined by you. Later you mention alpha. Do you understand the alpha is related to the confidence interval and confidence? Um, didn't I tell you to use a 95% confidence parameter? Didn't I define the interval upon which this confidence parameter would be based? Can we assume a log normal distribution? What more do YOU need to know, Bill?

I am not going to get into a statistics pissing match with you, Bill.

Finally, reagarding the 78 number, refute my statistics or don't. My guess is you will not.

Lurker

Lurker

13th August 2003, 01:56 PM

OK, Bill. I decided to make this REAL easy for you. When I said confidence I really meant 1-significance. Now do you understand the terms I was using? Does that make it any plainer to you?

You see, Confidence is directly related to Significance. So when you asked Thanz for a significance of 0.05, well why do you think I used 0.95 or 95% for my confidence oh so many posts ago? It seems YOU are the one who does not understand some things otherwise you would have caught that.

Sheesh.

Lurker

Lurker

13th August 2003, 02:15 PM

Originally posted by BillHoyt

Thanz and Lurker want us to believe they can come up with something. So far, that something is hot air, a drive-by assertion that 78 is "obviously inadequate".

Cheers,

Bill, I can't speak for Thanz but I came up with more than hot air. I provided a website where I told you what numbers to input 95% (or your 5% significance - a common value), the confidence interval +/-5% which was being most generous in my estimation, and the population size. It provided a value for sample size far different than 78.

Now, what have you proffered in return? You bleat about me not doing anything. You bandy stat terms about as if I am supposed to educate you on what they mean. You provide no refutation whatsoever. If you have a problem with my method, please enlighten me. I am fully aware their are problems with the method I used, perhaps you can find them? But I will note again that I provided a statistical method to test the 78 hypothesis. You have provided notthing in return. At least Claus had the common sense (I know how much you like that term) to not attack me saying I did nothing. You ignore what I have posted and continue saying I provided nothing.

Pot, meet kettle...

Lurker

Lurker

13th August 2003, 02:17 PM

Originally posted by Lurker

Statistical Test: http://www.surveysystem.com/sscalc.htm

Input 95% confidence in your data. Then a confidence interval of 5, which is VERY high for the multivariate statistics you are using here. With so many names, the +-5% for each name would be exceedingly high. Next, enter the population fo 280million. Have it calculate and voila! Out pops the number 384 for a sample size. Far higher than 78, right?

I agree it is not exact and leaves a bit to be desired but it is a decent approximation, no?

Let me bump this post again so Bill can address the specific problems in it.

Lurker

Lurker

13th August 2003, 02:22 PM

As voidx said to you, Bill: "...so rather than go out of you're way to stump them, why not explain it to us so we can move on."

Yes, Bill. Feel free to show me where I went wrong. Please use your 0.05 significance and a reasonable confidence interval to show me how 78 is a reasonable number.

I have provided evidence it is not. You may disagree with that evidence but thus far you have not said you do. Why do you think that is? Instead you start creating quizzes. How nice.

Lurker

AlienX

13th August 2003, 03:06 PM

I'm no statistician but i'm not certain this calculator is applicable here. This is for OPINIONS which CHANGE especially for elections and such - thus inherantly I would assume a larger population sample is required to try and account for this (it will be built into the calculation obviously - a variability factor). With things like names it's FIXED (well you can change your name - but your name is not an opinion) thus the confidence level of any poll would be significantly higher for names - when you do the poll on the same sample a week later it's still the same (well apart from the dpoll name changers) thus a lower than what you have calculated minimum sample value would obviously be the result.

So in this instance 78 COULD indeed be ample sample poulation - But i'm not sure of the exact value I am certain it's definitely lower than 384 when the correct applicable calculation is applied allowing for the fact that what is being sampled is fixed.

I suppose you could say on day 1 you could sample 1000 people then sample exactly the same 1000 people the next day and get a different result with respect to opinions - but their names would not have changed.

Am I alone in thinking that this calculation produces a higher number due to the possible variation in opinion being accounted for.

Anyone who knows stats better than us care to comment - i think what I have said is logical. Sorry Lurker I'm not taking sides or anything here it's just something which instantly struck me looking at the calculator.

AlienX

Thanz

13th August 2003, 03:27 PM

Well, I must say that I did not expect this kind of thread explosion on this point. Maybe I should have expected it from Bill Hoyt, given some of his other tangents. But I digress.

Anywhoo, I did not at any time claim to be a statistician, nor to have calulated any figures. As lurker has been keen to understand, I used a common sense approach to my thinking.

It is common sense to state that 78 letter guesses is not an adequate sample to tell us anything meaningful about the rate of guesses by JE and how they compare to the general population.

It is also common sense to say that a sample of 1000 is likely to be better than a sample of 78.

So far, Lurker is the only one to have come up with any statistical basis for a sample size, and it supports my assertion.

Bill Hoyt has simply made the bald statement that my assertion is wrong, with absolutely no back up at all.

So, Mr. Hoyt, are you going to back up your bare assertion, or not?

BillHoyt

13th August 2003, 04:09 PM

Lurker,

You seem upset I did not comment on your previous answer. I was hoping somebody else would, frankly, because, frankly, you will be far more upset when I am done commenting.

Here is what you said:

Experimental Design: Let us assume you can randomly choose the samples from the total population. By doing this, you are assuming a best possible scenario. Anything else would require a LARGER sample size to make up for the deficiency.

Statistical Test: http://www.surveysystem.com/sscalc.htm

Input 95% confidence in your data. Then a confidence interval of 5, which is VERY high for the multivariate statistics you are using here. With so many names, the +-5% for each name would be exceedingly high. Next, enter the population fo 280million. Have it calculate and voila! Out pops the number 384 for a sample size. Far higher than 78, right?

I agree it is not exact and leaves a bit to be desired but it is a decent approximation, no?

One of the first things you learn in statistics is to choose what is applicable. You have chosen:

o the wrong tool!
o the wrong pdf
o even the wrong space, for crying out loud
o to raise the issue of “multivariate statistics” when this has nothing to do with multivariate analysis

Let’s get down to brass tacks. The number of people in the population has nothing to do with anything! The number of letters in the alphabet is your basis space. There are, for the record, 26. Let me repeat that for you: who cares what the population size is? Nobody! Who cares about the number of names even. Both figure prominently in your discussion quoted above, and neither has anything to do with anything. All we need is a good frequency table with an entry for each letter. Period. End of story.

Now that means the pdf we need to examine here is in discrete space, not continuous space. But you are clearly in outer space, I’m afraid, and out of your league.

Now, about the tool; the “sample size calculator”. This is to see what sample size you need to properly represent your population. This is what you want in a survey. This is not what you want for hypothesis testing. You want to construct a null hypothesis, take your measurements and compare the results with what you should have gotten had your null hypothesis been correct.

Care to spin again, Lurker? And you wonder why I didn’t comment the first time around?

BillHoyt

13th August 2003, 04:17 PM

Originally posted by Thanz
So far, Lurker is the only one to have come up with any statistical basis for a sample size, and it supports my assertion.

Bill Hoyt has simply made the bald statement that my assertion is wrong, with absolutely no back up at all.

So, Mr. Hoyt, are you going to back up your bare assertion, or not?

Are you not reading my posts Thanz? Read them all. I have told you what questions you need to ask. The problem is, you have such a poor understanding of stats, you don't understand the meaning. With my last post, it should be abundantly clear that Lurker is not merely barking up the wrong tree or in the wrong forest. He's in the wrong space.

voidx

13th August 2003, 04:47 PM

But see, now we're getting somewhere. I'm sure Lurker has no problem admitting if he's done something wrong (if that's the case here, I don't know enough to tell). So I actually wish you had commented earlier, and in fact Lurker asked everyone to as he was sure it wasn't a perfect model. All you had to do was describe what you did just now, back then, and we could have avoided all the arm wrestling posts in between.

Loki

13th August 2003, 05:00 PM

Sheesh!!

My last formal encounter with statistics lies nearly 20 years in the past, so I ain't getting involved in the details of this...but surely the situation is :

1. We have a REAL statistics question here. We have a set of names, and a set of guesses. The data is available. The issue is whether the data is sufficient to apply a valid statistical analysis to, to derive a meaningful result.

2. Thanz says "78 seems too low". IMO he appears to be speaking from a "gut feel" or "commonsense" position (and for what it's wirth [not much] my first reactrion was to agree with this)

3. BillHoyt and Claus reply - "you can't say that for sure - what are the details of the statisical analysis you wish to run?". They are completely correct - Thanz can't make that statemetn with confidence unless he can fill in the details.

4. And this is where it all goes off the rails - THanz replies by asking "so do you think 78 is valid"? Bill and Claus reply with "do you understand why you can't say 78 is invalid?". And so on, and so on.

5. Lurker jumps in and offers a "possible" explanation for why 78 is too low.

6. Side issue develops where Lurker and Bill debate the applicability of Lurkers' analysis.

Look, all I want to know is if 78 is valid or not for the purpose it was raised!!

Bill and Claus - you're right! Thanz, whether he realises it or not (and I think he does!), is just expressing a "gut feeling" when he says 78 seems too low, and he hasn't backed that up.

Fine. But the real question Tnanz is/was asking (IMO) is "is 78 too low"? Or if he's not, then I am!!! Is anyone prepared to express an opinion in support of the validity of 78? If *no one* will offer a reason to expect 78 to have any possible validity, then what the hell is the point of all this? If Bill or Clause *do* have reason to believe it's valid, how about posting that so I can see it without having to spend the next two weeks on a statistics refresher course.

Anyone care to try and answer the real question, rather than just bash Thanz for making a statement he already knows was not a "certified fact"?

BillHoyt

13th August 2003, 06:51 PM

Originally posted by Loki
Sheesh!!
...
Anyone care to try and answer the real question, rather than just bash Thanz for making a statement he already knows was not a "certified fact"?
Loki,

Please re-read Thanz' last post. He says there that he never claimed expertise in statistics. He has yet to state his unqualified assertion was in error. I think, however, he is having a good think about it at the moment.

Did you read my post about this? The answer depends upon the question being asked. Before you even get into the statistics of it, though, you have to ask yourself: what question do you want to answer? To do that, you have to frame an hypothesis. The basic question is: are JE's guesses skewed or are they representative of the names out there. That is, does he guess "J" at about the frequency of all the "J" names out there. Let's say "J" names occur at p = .1. We'd expect to see 7.8 "J" guesses on chance alone then. Let's run with this bit alone for a bit.

I would frame the hypothesis that JE is cold-reading and his "J" guesses will show this by being significantly more frequent than chance alone would dictate. The next step is to turn this hypothesis on its head. This is called the "null hypothesis", and it would look like this: JE is not cold-reading and his "J" guesses will not be significantly higher than chance alone would dicatate Then I would select the significance level, probably .05 and look up the one-tailed test for significance. Why one-tailed? Because of how I framed the null hypothesis: I'm only interested in how unlikely it is to have had this high a number assuming his guesses were really random.

Now I'm assuming that J is the most frequent first initial. I would actually choose that from the control data; the names database. I am also assuming I would simply look at one such datum. There are, of course, several initials I could test in this way. The problem of data mining comes into play, though, and I really must set my significance level higher if I want to do this.

The next problem that comes into play is making certain that I set these tests up a priori!

Now the third problem is yet another salient point that should signal us how clearly wrong the bald assertion that n=78 is inadequate is! Let us say we want to know whether a coin is fair and toss it 78 times. Is this enough? Maybe yes, maybe no. Let's say we get 77 heads and 1 tail. I'd say we don't need to look at a set of tables to know that is not a fair coin. Let's say we get 34 heads and 33 tails. That, I'd say is probably fair, but I'd have to look at the tables to be sure. This underscores the final blow to the bald assertion: it also depends on the data!

Cheers,

Loki

13th August 2003, 08:06 PM

BillHoyt,

Thanks! Surprising enough, that's all pretty clear. So I guess the only thing to ask is "is 78 a valid sample" if :

The null hypothesis is : "JE is not cold-reading and his "J" guesses will not be significantly higher than chance alone would dicatate";

The significance level we seek is .05.

The data is available - 78 sample letters, and 8000 (approx) in the original names database ;

I could probably (eventually) do the math, but since you seem to be on top of this (and I'm feeling lazy!!), do you have a "yes/No" answer?

Please re-read Thanz' last post. He says there that he never claimed expertise in statistics. He has yet to state his unqualified assertion was in error. I think, however, he is having a good think about it at the moment.
It's a dangerous game to try and state what someone else is thinking (unless you're JE!), but it seems to me that if Claus (originally) and you had taken a slightly less aggressive approach in pointing out Thanz's "unwarranted assumption" that he'd have probably agreed several pages ago that "78 is too low" was a statement he couldn't support. It looked to me like Thanz wandered into an area that, in hindsight, he'd probably wished he hadn't, and you guys rushed in with all guns blazing.

Your point is well made, and worth making, but I'm not sure that Thanz was really arguing against you. If I had to try and paraphrase Thanz's stance it would be "My instinct says that 78 is too low - can anyone say why it isn't too low?". You've demonstrated perfectly well that "instinct" really isn't reliable in this situation. What you haven't shown is that, in fact, instinct - in this specific instance - is wrong. There's every chance that, even if Thanz has used the unreliable tool of "commonsense" to arrive at an answer ("78 is too low"), he may in fact be right!

Kerberos

14th August 2003, 12:30 AM

Originally posted by Loki
BillHoyt,

Thanks! Surprising enough, that's all pretty clear. So I guess the only thing to ask is "is 78 a valid sample" if :

The null hypothesis is : "JE is not cold-reading and his "J" guesses will not be significantly higher than chance alone would dicatate";

The significance level we seek is .05.

The data is available - 78 sample letters, and 8000 (approx) in the original names database ;

I could probably (eventually) do the math, but since you seem to be on top of this (and I'm feeling lazy!!), do you have a "yes/No" answer?

Well I can't calculate the reliability of the original name database since I don't know the math and since it's not random and not a US database, but the calculation of statistical significance for 78 letters (n) and 4 hits (r) or less with a 0.1454 chance of a hit (p) is fairly simple.
It's called something like binomial distribution (my book is in Danish and it’s called binomialfordeling in case Claus knows the proper English name).

The formula goes P(X=r) = n!/(r!(n-r)!*p^r*(1-p)^(n-r). The chance of getting 4 hits or less is then P(X=4) + P(X=3) + P(X=2) +P(X=1) + P(X=0) = 0.008 assuming the names come at random, but in a national broadcast on CNN I don't think this is an unreasonable assumption.

Of course I also used two snippets of edited transcripts. Without these there'd be 73 guesses and 3 hits which would occur 0.004 of the time by chance alone. So assuming that the name-database is accurate (which I admit is debatable) then yes we do achieve a significance level of less than 0.05.

Clancie

14th August 2003, 12:39 AM

Some posts have referred to JE getting names and some to initials. You are going to calculate them separately, aren't you, Bill?

After all, the odds of getting "Alf" (as he did yesterday) should be quite different than just saying "Is there an 'A' name?"

Also, OT but of interest, yesterday JE did the same reading (two people together) for basically the entire time (including their post analysis, which was pretty short and excluding the brief intro and conclusion, the two sitters' reading made up all the rest of the show).

BillHoyt

14th August 2003, 04:27 AM

Originally posted by Kerberos
Well I can't calculate the reliability of the original name database since I don't know the math and since it's not random and not a US database, but the calculation of statistical significance for 78 letters (n) and 4 hits (r) or less with a 0.1454 chance of a hit (p) is fairly simple.
It's called something like binomial distribution (my book is in Danish and it’s called binomialfordeling in case Claus knows the proper English name).
The UK database has two problems that I can see: first, is its representativeness of the UK population at large and second is its representativeness of the US population. Because of those problems, I think the best solution is to get a good US database, such as the Social Security one somebody had mentioned earler.
The formula goes P(X=r) = n!/(r!(n-r)!*p^r*(1-p)^(n-r). The chance of getting 4 hits or less is then P(X=4) + P(X=3) + P(X=2) +P(X=1) + P(X=0) = 0.008 assuming the names come at random, but in a national broadcast on CNN I don't think this is an unreasonable assumption.
I'm not sure I understand why you calculated "four hits or less"? I'd be looking for a one-tailed test of "x hits or more". Could you explain this more?

Cheers,

BillHoyt

14th August 2003, 04:30 AM

juninho

14th August 2003, 05:03 AM

Originally posted by BillHoyt

For this hypothesis, an "Alf" would be an "A". You can't switch statistics in the middle of the stream.

Cheers,

Yes, entirely and when it isn't Alf first time he would go on to or "Andy" or "Albert" or something. So I think it is perfectly acceptable to count it is an "Initial" guess.

Plus, anyway, we're trying to come up with something that we can test - we can't possibly know how many Alfs, Susans or Garys any one individual would know or "it means something to". Impossible to test.

Therefore, let's just stick to the initial of the forename.

I think the best way to go about analysis, considering we've got a relatively small sample size, is to lump letters together and work out the likelihood of a particular group of letters occuring.

I agree that you should go for some US data and I'd probably lean towards the census data as (someone pointed out) it includes the immigrant population who are just as likely to visit one of JE's shows as anyone else.

Anway, thanks Clancie, we've just increased the sample size by one. That A for Alf will now make the sample size 79.

BTW, I've got no problem including any of the crossing over stuff.

Lurker

14th August 2003, 05:18 AM

Originally posted by BillHoyt
Lurker,

You seem upset I did not comment on your previous answer. I was hoping somebody else would, frankly, because, frankly, you will be far more upset when I am done commenting.

Here is what you said:

One of the first things you learn in statistics is to choose what is applicable. You have chosen:

o the wrong tool!
o the wrong pdf
o even the wrong space, for crying out loud
o to raise the issue of “multivariate statistics” when this has nothing to do with multivariate analysis

Let’s get down to brass tacks. The number of people in the population has nothing to do with anything!

Um, I never said it did. Since the calculator asked for it I provided IT WOULD matter if the sample size were close to the size of the population but I agree that the population is unimportant for our analysis. But this would not always be the case. But you knew that, right Bill?

The number of letters in the alphabet is your basis space. There are, for the record, 26. Let me repeat that for you: who cares what the population size is? Nobody! Who cares about the number of names even. Both figure prominently in your discussion quoted above, and neither has anything to do with anything. All we need is a good frequency table with an entry for each letter. Period. End of story.

Agreed.

Now, about the tool; the “sample size calculator”. This is to see what sample size you need to properly represent your population. This is what you want in a survey. This is not what you want for hypothesis testing. You want to construct a null hypothesis, take your measurements and compare the results with what you should have gotten had your null hypothesis been correct.

All right, Bill. So how does it not relate? Aren't we trying to figure out if each mean is properly represented? Each mean being the frequency each letter is represented.

I note you did not seem interested in addressing your lack of knowledge of confidence levels. You bleated on about significance as if that were vastly different than confidence. Why are you ignoring that? Perhaps you are not as good at statistics as you profess.

Regardless, I stand by my assertion that 78 is probably too low a number. I would welcome any REAL analysis though.

Lurker

Lurker

14th August 2003, 05:26 AM

Originally posted by BillHoyt

Now the third problem is yet another salient point that should signal us how clearly wrong the bald assertion that n=78 is inadequate is! Let us say we want to know whether a coin is fair and toss it 78 times. Is this enough? Maybe yes, maybe no. Let's say we get 77 heads and 1 tail. I'd say we don't need to look at a set of tables to know that is not a fair coin. Let's say we get 34 heads and 33 tails. That, I'd say is probably fair, but I'd have to look at the tables to be sure. This underscores the final blow to the bald assertion: it also depends on the data!

Cheers,

Well, it DOES depend on the data. Your analogy above is the opposite of great. First off, you did not bother to provide the analysis. Prove that the coin is invalid. Otherwise you are using the same method Thanz and I used - common sense. :) (I think I better start using smilies again cause some people take what I say far too seriously.)

Also, you only have two possible outcomes to map the frequency of. Our analysis has 26.

Let us assume the probability of all 26 outcomes is the same ~4%. Clearly we would need AT MINIMUM 26 samples to see that. But what confidence would we have in the data? Oh, I am sorry. Bill didn't understand that, allow me to rephrase What significance does the data have?

Lurker

Lurker

14th August 2003, 05:34 AM

Originally posted by Lurker

Let us assume the probability of all 26 outcomes is the same ~4%. Clearly we would need AT MINIMUM 26 samples to see that. But what confidence would we have in the data? Oh, I am sorry. Bill didn't understand that, allow me to rephrase What significance does the data have?

Lurker

Uh, oh. Now I am answering myself. Not a good sign. I calculated the probability of in 26 tries getting all 26 letters. Not good at 0.0000000000655. Clearly we will need more than 26 guesses in this case as otherwise we will erroneeously think a letter is 0%.

Lurker

Lurker

14th August 2003, 06:02 AM

Lurker

14th August 2003, 06:06 AM

Further, I was not aware the null hypothesis was JE was cold reading or not. I thought we were trying to determine if 78 random names would be able to properly represent (assuming a specific degree of confidence in a specific confidence interval) the frequencies of the first letters of the first names. A histogram, if you will.

Lurker

Kerberos

14th August 2003, 06:33 AM

Originally posted by BillHoyt

I'm not sure I understand why you calculated "four hits or less"? I'd be looking for a one-tailed test of "x hits or more". Could you explain this more?

Cheers,

Well the question is if JE avoids rare letters (hits), so what I'm calculating is the probability that JE would by chance alone get as few rare letters or less. I could also calculate the probability of getting more hits, but that just 1-(probability of 4 hits or less), but calculating 4 or more wouldn't make sense.

Originally posted by juninho

I agree that you should go for some US data and I'd probably lean towards the census data as (someone pointed out) it includes the immigrant population who are just as likely to visit one of JE's shows as anyone else.

Plus the record of the 1000 most popular names obviously doesn't include the least common names, and I suspect that rare letters will be overrepresented in the least common names.

I confirmed this suspicion, as far as it can be confirmed using the data I had, by finding that X, U, Q, Y, Z, O, I, V, F and N were the 10 least common letters used and were the initial letter of 6,78% of the total names, but if I only counted the names that less than 100 people had, then these 10 letters were used 13,02% of the time as the initial letter.

The problem with the Census data is that I can't copy it into Excel, order it by first letter and count it the same way I can with the other source, because both the names and the numbers get posted into the same box and I don't know how to separate them.

If somebody with greater knowledge of Excel or some other appropriate program could do it that would be great, but I can't and doing it manually would take a long time and I'd rather see if somebody with greater knowledge of computers could do it automatically.

Originally posted by Lurker

you only have two possible outcomes to map the frequency of. Our analysis has 26.

Actually our analysis only has two outcomes - rare and common letters. We define fx the 5, 10 or 15 rarest letters as rare and the rest as common, and then we count the number of guesses that fall into each category. Then we can use the formula I posted before to calculate the probability that this is a coincidence.

Calculating the chance that the letters would be distributed exactly as they are might be interesting, but I don't know how to do that and as you pointed out we don't have enough data, so we reduce the problem to a binary form.

CFLarsen

14th August 2003, 06:42 AM

Kerberos,

No big deal for me.

I'll strip the list any way you want it. Email me.

Kerberos

14th August 2003, 06:53 AM

Originally posted by Lurker

Sorry to interrupt. Where did the 0.1454 come from?

thanks!

Lurker

That's the frequency with which the letters Z, X, Y, U, O, Q, I, V, W, T, N and F appear in the sample data from juninho. They appear 19 + 14 + 72 + 9 + 4 + 42 + 85 + 186 + 214 + 239 + 177 + 146 = 1207 times out of 8303 names 1207/8303 =0,1454 or 14,54%

Lurker

14th August 2003, 07:15 AM

Originally posted by Kerberos

Actually our analysis only has two outcomes - rare and common letters. We define fx the 5, 10 or 15 rarest letters as rare and the rest as common, and then we count the number of guesses that fall into each category. Then we can use the formula I posted before to calculate the probability that this is a coincidence.

Calculating the chance that the letters would be distributed exactly as they are might be interesting, but I don't know how to do that and as you pointed out we don't have enough data, so we reduce the problem to a binary form.

Ah, I was operating under the assumption we were attempting to obtain data which would determine the frequency of each letter (why I mentioned 26 possiblities a couple times). If that is not the case, then perhaps 78 IS a reasonable number, depending on how broadly you are creating the groups.

I also concur that creating a binary problem makes it much more tractable.

Lurker

Lurker

14th August 2003, 07:21 AM

Bill Hoyt:

I am not sure, but perhaps you and I were talking about two different problems. If the problem was:

1. Let's determine the frequency of each letter, I stand by my assertion that 78 is far too small a sample size to adequately define each letter's frequency.

2. Let's arbitrarily group the letters into common and rare, then 78 probably IS an adequate sample size.

So, until I hear from you on which problem YOU thought the 78 was referring to, I will turn the flamethrower *off*.

Lurker

BillHoyt

14th August 2003, 07:21 AM

Originally posted by Lurker

Ah, I was operating under the assumption we were attempting to obtain data which would determine the frequency of each letter (why I mentioned 26 possiblities a couple times). If that is not the case, then perhaps 78 IS a reasonable number, depending on how broadly you are creating the groups.

I also concur that creating a binary problem makes it much more tractable.

Lurker

Great to hear, Lurker.

Cheers,

BillHoyt

14th August 2003, 07:31 AM

Originally posted by Kerberos
Well the question is if JE avoids rare letters (hits), so what I'm calculating is the probability that JE would by chance alone get as few rare letters or less. I could also calculate the probability of getting more hits, but that just 1-(probability of 4 hits or less), but calculating 4 or more wouldn't make sense.

I agree that we have two ways to see if JE's forename initials merely reflect the population or suggest cold-reading. The basic question is: does he concentrate on common initials at the expense of the rare. And I agree that we should test the rare initials for the low tail. We should also test the common initials on the high tail. However, depending on how many times we dip in the data well, we have to be careful of the significance level we use. (It also depends on the results, of course. If we test 20 times at p=.05, we expect to get 1 null hypothesis rejection on chance alone, and shouldn't think it is very meaningful. If we test 20 times a p=.05 and get 5 null hypothesis rejections, then we've got something.

I think also of interest might be overlaid histograms. The background would be the initial frequencies in the population at large, normalized to 78. The foreground would be the initial frequencies from the JE readings. I'd expect this graphic to show a fairly clear pattern of skewing in favor of common initials if JE is cold-reading.

Therein lies a problem, though, for clumping the edited CO shows together with full, live transcripts. We complicate the picture by adding more hypotheses with the CO shows. Warm-reading? Hot-reading? Adding the CO shows increasing the power of the analysis, but introduces these other complications.

Cheers,

BillHoyt

14th August 2003, 07:52 AM

Originally posted by Lurker
Bill Hoyt:

I am not sure, but perhaps you and I were talking about two different problems. If the problem was:

1. Let's determine the frequency of each letter, I stand by my assertion that 78 is far too small a sample size to adequately define each letter's frequency.

2. Let's arbitrarily group the letters into common and rare, then 78 probably IS an adequate sample size.

So, until I hear from you on which problem YOU thought the 78 was referring to, I will turn the flamethrower *off*.

Lurker

Lurker,

I've always seen the problem the way I stated it earlier:

I would frame the hypothesis that JE is cold-reading and his "J" guesses will show this by being significantly more frequent than chance alone would dictate. The next step is to turn this hypothesis on its head. This is called the "null hypothesis", and it would look like this: JE is not cold-reading and his "J" guesses will not be significantly higher than chance alone would dicatate .

Testing the rare letters for being "too low" is another approach because to favor the common letters would force a reduction in rare letter occurrence.

Cheers,

BillHoyt

14th August 2003, 08:19 AM

"dicatate"? Did I really write "dicatate"? Sheesh.

juninho

14th August 2003, 10:54 AM

Originally posted by CFLarsen
Kerberos,

No big deal for me.

I'll strip the list any way you want it. Email me.

OK, the US 1990 Census Data has now been done. I've adjusted the frequencies to take account of the fact that the Male Cumulative and Female Cumulative frequencies fell short of 100% (evidently because names are just too rare and they needed a cut off point with there stats or else the list would be too long). Anyway, for Females it is:

INITIAL SumOfadj freq
A 7.5358
B 5.1153
C 7.7746
D 6.4194
E 5.4074
F 1.3907
G 2.4782
H 2.1839
I 1.0364
J 9.0842
K 5.2919
L 7.6602
M 12.3456
N 2.3849
O 0.4421
P 3.4757
Q 0.0189
R 4.6177
S 7.8146
T 3.7246
U 0.0333
V 2.2383
W 0.8887
X 0.0089
Y 0.4343
Z 0.1011

For Males:

Initial SumOfadj freq
A 5.4365
B 4.1593
C 6.6470
D 8.4207
E 3.7916
F 2.1080
G 4.0737
H 2.3234
I 0.4220
J 17.6388
K 2.9054
L 3.5673
M 7.7299
N 1.2639
O 0.4631
P 2.7788
Q 0.0533
R 11.3883
S 4.5291
T 4.7235
U 0.0111
V 0.7263
W 4.6824
X 0.0144
Y 0.0144
Z 0.1410

And together (which is presuming that the sexes are equal and ignoring the slight difference in cumulative frequency but adjustments can be made if anyone can find the split):

Initial Totaladjfreq
A 6.4861
B 4.6373
C 7.2108
D 7.4201
E 4.5995
F 1.7493
G 3.2760
H 2.2536
I 0.7292
J 13.3615
K 4.0986
L 5.6137
M 10.0377
N 1.8244
O 0.4526
P 3.1273
Q 0.0361
R 8.0030
S 6.1718
T 4.2240
U 0.0222
V 1.4823
W 2.7855
X 0.0117
Y 0.2244
Z 0.1211

Enjoy. I'm going home.

jim_scotti

14th August 2003, 11:09 AM

Originally posted by Clancie
Some posts have referred to JE getting names and some to initials. You are going to calculate them separately, aren't you, Bill?

After all, the odds of getting "Alf" (as he did yesterday) should be quite different than just saying "Is there an 'A' name?"

The odds are irrelevant if he has been warm reading the audience prior to the taping. Since eyewitnesses have verified that JE and his crew spy on his audiences before taping withg live microphones and cameras innocently running beforehand, all bets are off on his getting these kinds of hits only through the use of his supposed psychic powers.

Jim.

neofight

14th August 2003, 11:35 AM

Originally posted by jim_scotti

The odds are irrelevant if he has been warm reading the audience prior to the taping. Since eyewitnesses have verified that JE and his crew spy on his audiences before taping withg live microphones and cameras innocently running beforehand, all bets are off on his getting these kinds of hits only through the use of his supposed psychic powers.

Jim.

Well, Jim, first of all, I think you mean "hot" reading, and not "warm" reading, and I would also suggest that you are mistaken about your eyewitnesses, because what you are asserting simply doesn't happen. BTW, Who are these "eyewitnesses" you speak of?......neo

CFLarsen

14th August 2003, 11:38 AM

renata

14th August 2003, 11:43 AM

Originally posted by CFLarsen

You got a hell of a nerve, asking people for references, when you rarely give any yourself!! :mad:

LOL! Claus, silly, the references are in private messages! You would not want confidences to be broken, would ya? ;)

Thanz

14th August 2003, 11:58 AM

Well. a lot has happened in this thread, and instead of addressing each post I will just give a general reply and hope I haven't left anything out.

First, as is obvious (I think) I did not do any calculations to come up with my statement that 78 is inadequate. I used a common sense approach, and figured that with 26 possible guesses, 78 is not enough to get an accurate distribution across the whole set, which is what I perceive as "meaningful".

Next, I admit that I did stall both CFLarsen and BillHoyt. With my previous acrimonious histroy with these two particular posters, I thought that they might just be trying to take me on for the sake of taking me on. Which is why I wanted at least one of them to commit to saying that 78 is adequate, as I thought (and still do believe) that this number is too small a sample. I apologize for the shenanigans.

As to the substance, I don't think that it is adequate to simply look at the mentions of one letter. I think what is required is a distribution of guesses across all possible outcomes (letters) and a comparison to the frequency in the normal population. And I still think that 78 is too small for such an analysis.

Further, we need to look at the source of the 78 and see if we are comfortable with it as representative of the guesses he makes. We only have data from LKL, and I think one CO. CO data is problematic as it will probably over emphasize hits. LKL data is problematic as it would seem that that is the only place where he does these more rapid readings. If we had some data from seminars, for example, I think it would be much better. I am not convinced that 70-odd guesses from LKL are representative of JE readings in general.

I am fully aware that this data is hard to come by, and that sometimes you have to make do with what you have got. But you always have to recognize the problems that you have with the data and if appropriate, admit that the data simply isn't good enough.

One other point - I don't think that a guess of "John" can be equated with a guess of "J". The probability of getting a "J" is higher than getting a "John", which makes them different guesses. If "John" is a miss, and he goes on to "some other J name" then I would agree. I don't think that this is changing the statistics - I think it is making sure we get the right statistics in the first place.

If I missed anything, let me know and I'll try to answer.

CFLarsen

14th August 2003, 11:58 AM

Originally posted by renata
LOL! Claus, silly, the references are in private messages! You would not want confidences to be broken, would ya? ;)
Not silly at all. Neo referred to posts on TVTalkshows. It was only later she changed it to private messages.

Silly! :)

jim_scotti

14th August 2003, 12:02 PM

Originally posted by neofight

Well, Jim, first of all, I think you mean "hot" reading, and not "warm" reading, and I would also suggest that you are mistaken about your eyewitnesses, because what you are asserting simply doesn't happen. BTW, Who are these "eyewitnesses" you speak of?......neo

You're right, it is "Hot reading" I was talking about. The eyewitness is one Michael O'Neill who was quoted in a Time Magazine article about his experiences as an audience member on CO. Here is an interesting article about Edward that includes quotes from O'Neill. http://www.skeptic.com/newsworthy13.html

JIm.

renata

14th August 2003, 12:03 PM

Originally posted by CFLarsen

Not silly at all. Neo referred to posts on TVTalkshows. It was only later she changed it to private messages.

Silly! :)

Erggh-- I meant references Neo was demanding from Jim_Scotii are in private messages and could not be revealed...I was teasing because Neo evaded substantiating her own claims about Instig8r!

Nothing sucks more than explaining a poorly crafted joke

*mope*:(

Thanz

14th August 2003, 12:10 PM

Originally posted by renata

Erggh-- I meant references Neo was demanding from Jim_Scotii are in private messages and could not be revealed...I was teasing because Neo evaded substantiating her own claims about Instig8r!

Nothing sucks more than explaining a poorly crafted joke

*mope*:(
If it makes you feel any better, I got the joke.

Have a head rub, mopey. :rub:

renata

14th August 2003, 12:18 PM

Originally posted by Thanz

If it makes you feel any better, I got the joke.

Have a head rub, mopey. :rub:

Ahhh all better, thanks! :D

We shall just write this one off to dumb humorless Danes, OK? ;)

BillHoyt

14th August 2003, 12:22 PM

Originally posted by Thanz
First, as is obvious (I think) I did not do any calculations to come up with my statement that 78 is inadequate. I used a common sense approach, and figured that with 26 possible guesses, 78 is not enough to get an accurate distribution across the whole set, which is what I perceive as "meaningful".
I have no idea why you think you need an accurate distribution. Please read my previous posts on what the hypotheses of interest are.
As to the substance, I don't think that it is adequate to simply look at the mentions of one letter. I think what is required is a distribution of guesses across all possible outcomes (letters) and a comparison to the frequency in the normal population. And I still think that 78 is too small for such an analysis.
Justify this by stating the hypothesis and null hypothesis, please.
Further, we need to look at the source of the 78 and see if we are comfortable with it as representative of the guesses he makes. We only have data from LKL, and I think one CO. CO data is problematic as it will probably over emphasize hits. LKL data is problematic as it would seem that that is the only place where he does these more rapid readings. If we had some data from seminars, for example, I think it would be much better. I am not convinced that 70-odd guesses from LKL are representative of JE readings in general.
We don't care whether they are representative of JE guesses. We only care whether they appear to be cold reading or otherwise.
I am fully aware that this data is hard to come by, and that sometimes you have to make do with what you have got. But you always have to recognize the problems that you have with the data and if appropriate, admit that the data simply isn't good enough.
You still need to demonstrate the data are not good enough.
One other point - I don't think that a guess of "John" can be equated with a guess of "J". The probability of getting a "J" is higher than getting a "John", which makes them different guesses. If "John" is a miss, and he goes on to "some other J name" then I would agree. I don't think that this is changing the statistics - I think it is making sure we get the right statistics in the first place.
Malarky. The probability of finding an orange on the ground in Florida is higher than the probability of finding an apple. That doesn't mean I can't design my experiment to collect statistics on the number of fallen fruit, regardless of variety.

Cheers,

jim_scotti

14th August 2003, 12:25 PM

Originally posted by Thanz

If it makes you feel any better, I got the joke.

Have a head rub, mopey. :rub:

I got the joke too, even if there wasn't any private communication involved..... :hit: Not everyone on this list needs joke explainations....

Jim.

Thanz

14th August 2003, 12:42 PM

Originally posted by BillHoyt

I have no idea why you think you need an accurate distribution. Please read my previous posts on what the hypotheses of interest are.

Justify this by stating the hypothesis and null hypothesis, please.
If JE is cold reading, then the distribution of his guesses on the starting letters of first names will be skewed in favour of the most prevalant first name letters in the general population. Conversely, if he is not cold reading, the distribution of letter guesses will be close to the normal distribution in the population.

Not perfect, as the people who get read by JE are not truly random, and may not represent the whole of the population. I think it is probably impossible to get name data for just those who are within the group that may be read by JE however.

We don't care whether they are representative of JE guesses. We only care whether they appear to be cold reading or otherwise.
But we can only tell if he is cold reading if the guesses ARE representative of his guesses overall. Otherwise, you cannot say that he is saying "J" because it is common or "J" because that is what is "coming through" to him. If they are not representative of his guesses as a whole, then you could just be proving that those people who got read on LKL knew more people with common names, rather than anything to do with JE and cold reading.

If we expand the sample size, the risks of this are diminished.

You still need to demonstrate the data are not good enough.
I explained my problems with the data already. I think that you need a whole distribution, not just one letter. I think that 70 odd guesses from LKL are not necessarily representative. I think that CO data, as edited, is suspect.

Malarky. The probability of finding an orange on the ground in Florida is higher than the probability of finding an apple. That doesn't mean I can't design my experiment to collect statistics on the number of fallen fruit, regardless of variety.
That is not an accurate analogy. You are assuming that all fallen fruit are equal. If we want to know how many oranges there are, we don't include apples.

We are trying to see if he is cold reading. When doing a reading, a guess of a specific name is not the same as a guess of a letter. They are not both fruits, in your parlance.

BillHoyt

14th August 2003, 01:15 PM

Originally posted by Thanz
If JE is cold reading, then the distribution of his guesses on the starting letters of first names will be skewed in favour of the most prevalant first name letters in the general population. Conversely, if he is not cold reading, the distribution of letter guesses will be close to the normal distribution in the population.
This has been stated several times before.

But we can only tell if he is cold reading if the guesses ARE representative of his guesses overall. Otherwise, you cannot say that he is saying "J" because it is common or "J" because that is what is "coming through" to him. If they are not representative of his guesses as a whole, then you could just be proving that those people who got read on LKL knew more people with common names, rather than anything to do with JE and cold reading.
Not true. Moreover, you are now requiring we seek data we will never, ever be able to get: All the JE sitters, plus all possible people associated with them. Come on. Get real.

If we expand the sample size, the risks of this are diminished.
Yes.

That is not an accurate analogy. You are assuming that all fallen fruit are equal. If we want to know how many oranges there are, we don't include apples.

We are trying to see if he is cold reading. When doing a reading, a guess of a specific name is not the same as a guess of a letter. They are not both fruits, in your parlance.
Give it up, Thanz. You have no idea what you are talking about! You do not understand the hypothesis being tested and you do not understand how to test that hypothesis. Furthermore, this attempt to disallow "James" as a "J" is hysterical!

neofight

14th August 2003, 01:16 PM

Originally posted by jim_scotti

You're right, it is "Hot reading" I was talking about. The eyewitness is one Michael O'Neill who was quoted in a Time Magazine article about his experiences as an audience member on CO. Here is an interesting article about Edward that includes quotes from O'Neill. http://www.skeptic.com/newsworthy13.html

JIm.

Hi, Jim. I see that you are new here, so I understand that you might not have read all the discussions we've had here about Michael O'Neill or about Shermer's article that you reference.

However, the truth is that everything that O'Neill is quoted as having said, was mere speculation on his part, and has no basis in fact. Neither was there any sort of investigative reporting done on his story whatsoever. James Randi and Michael Shermer simply took O'Neill at his word, and printed it as fact, even though he had no proof of any of his allegations.

It was a detestable bit of shabby journalism, Jim, and Leon Jaroff, Michael Shermer and James Randi ought to be ashamed. They set a new low in journalism standards with these articles.

If you want to do a search, you could probably find in what old threads this has been discussed.......neo

Clancie

14th August 2003, 01:27 PM

Posted by Bill Hoyt

Thanz: When doing a reading, a guess of a specific name is not the same as a guess of a letter.

Bill Hoyt: ...This attempt to disallow "James" as a "J" is hysterical!
So...Bill. Do you consider JE getting a hit with "I'm getting a 'J' name" to be the same as JE getting a hit saying, "I'm getting 'James'."

CFLarsen

14th August 2003, 01:30 PM

Originally posted by neofight
However, the truth is that everything that O'Neill is quoted as having said, was mere speculation on his part, and has no basis in fact. Neither was there any sort of investigative reporting done on his story whatsoever. James Randi and Michael Shermer simply took O'Neill at his word, and printed it as fact, even though he had no proof of any of his allegations.

Whoa....aren't you similarly taking the word of each and every sitter on CO, at the seminars and at private readings, too?

Why is this different, neo? Because O'Neill is critical of JE, right?

Originally posted by neofight
It was a detestable bit of shabby journalism, Jim, and Leon Jaroff, Michael Shermer and James Randi ought to be ashamed. They set a new low in journalism standards with these articles.

You accept the word of every sitter who validates that JE is a real medium. What does that make you, neo?

Originally posted by neofight
If you want to do a search, you could probably find in what old threads this has been discussed.......neo

Or, of course, you could show him....nah, of course you couldn't....

Darat

14th August 2003, 01:36 PM

Originally posted by Clancie

So...Bill. Do you consider JE getting a hit with "I'm getting a 'J' name" to be the same as JE getting a hit saying, "I'm getting 'James'." [/B]

Answering for him, but in the context of what is being discussed the answer is "yes", unless you want to claim that "James" doesn't start with a "J"!

Kerberos

14th August 2003, 01:45 PM

Originally posted by BillHoyt

I agree that we have two ways to see if JE's forename initials merely reflect the population or suggest cold-reading. The basic question is: does he concentrate on common initials at the expense of the rare. And I agree that we should test the rare initials for the low tail. We should also test the common initials on the high tail.

True but I did the calculation on my pocket calculator, and it's far easier to do such calculations when the number of hits is small. It's probably possible to calculate a scenario with relatively many hits quickly in excel, but I haven't looked at that yet.

Originally posted by BillHoyt

Therein lies a problem, though, for clumping the edited CO shows together with full, live transcripts. We complicate the picture by adding more hypotheses with the CO shows. Warm-reading? Hot-reading? Adding the CO shows increasing the power of the analysis, but introduces these other complications.

Cheers,

Well the way to go in my opinion would be to analyze the CO transcripts and see if a pattern emerges. If it does then it indicates cold reading and if it doesn't then it could be because of editing, warm reading, hot reading or genuine mediumship.

jim_scotti

14th August 2003, 01:45 PM

Originally posted by neofight

However, the truth is that everything that O'Neill is quoted as having said, was mere speculation on his part, and has no basis in fact. Neither was there any sort of investigative reporting done on his story whatsoever. James Randi and Michael Shermer simply took O'Neill at his word, and printed it as fact, even though he had no proof of any of his allegations.

So I suppose it's just "mere speculation" when someone is critical of your hero, yet when a sitter on CO credulously supports JE, it is solid fact? While O'Neill might not have had proof, he sure had solid observations of what was going on and it doesn't take much circumstantial evidence to show that hot reading is very possible under those conditions.

Originally posted by neofight

It was a detestable bit of shabby journalism, Jim, and Leon Jaroff, Michael Shermer and James Randi ought to be ashamed. They set a new low in journalism standards with these articles.

Is it detestable because it is critical of JE or do you actually have some real specific criticism of the report? It sure seems a lot more reputable than the unskeptical hogwash coming from the psychics who claim to be talking to the dead and their uncritical supporters.

Originally posted by neofight

If you want to do a search, you could probably find in what old threads this has been discussed.......neo

Well thanks, I see you are good at supplying references for me.

Jim.

neofight

14th August 2003, 01:50 PM

quote:
--------------------------------------------------------------------------------
Originally posted by renata
LOL! Claus, silly, the references are in private messages! You would not want confidences to be broken, would ya?
--------------------------------------------------------------------------------

Originally posted by CFLarsen

Not silly at all. Neo referred to posts on TVTalkshows. It was only later she changed it to private messages.

Silly! :)

Boy, what a couple of butt-in-skis you two are. :rolleyes:

Claus, you, sir, are a liar, as per usual. Nothing new here. You excell at misrepresenting what others say, and by now, others know it too, believers and skeptics alike.

I never referred to "posts" on TVTalkShows in the context that you say I did. When I mentioned the "posts" it was in reference to the hateful language that Instig8R habitually used when discussing JE.

Other than that, the only posts mentioned that referred to this issue were, as Instig8R herself pointed out, posts written by atmytv and Steve Grenard, commenting on her JE bashing. Atmytv, especially, found her remarks to be offensive.

Other posts, by RC and Celter, supported Instig8R, and those I would feel free to discuss with her, since they were public. I even re-posted atmytv's post here, to show exactly what he had said about it.

Anything written to me privately, either by PM or e-mail, I am under no obligation to quote, and neither do I need it as evidence, since there were three individuals, including myself, who had already made public comments concerning Instig8R's nasty JE references.

So for once, Claus, why don't you try to get something right, as long as you are going to butt into other people's business anyhow?.....neo

Thanz

14th August 2003, 02:00 PM

Originally posted by BillHoyt
Not true. Moreover, you are now requiring we seek data we will never, ever be able to get: All the JE sitters, plus all possible people associated with them. Come on. Get real.
I am NOT requiring this, not at all! I even acknowledge in my post that getting this kind of data is impossible.

What I am saying is that we want a reasonably representative sample of his guesses, so that the chances of the correlation being with the sitters rather than JE's guesses is diminished. We can't take data from one reading and say that since he read 10 people, and 3 had J names, he was cold reading. There is, IMO, too great a chance that this is due to the sitters rather than JE. If, however, we had a larger, representative sample of how JE operates, and he still guessed J at a 30% rate, we have a much more meaningful study.

Give it up, Thanz. You have no idea what you are talking about! You do not understand the hypothesis being tested and you do not understand how to test that hypothesis. Furthermore, this attempt to disallow "James" as a "J" is hysterical!
I will use extremes to illustrate my point. Let's say we have two purported mediums, JE and SB. And let's also say we look at the technique of both, and JE always guesses specific names, while SB only ever guesses initials. They get the same number of hits. Who do you think it is more likely to be cold reading? Isn't the case against SB much stronger, as everytime she takes a guess she casts a much wider net? Isn't guessing "James" more risky than guessing "J"? Don't you think that this should be taken into account? Your method would equate both JE and SB to be doing the same thing, when I think it is clear they are not.

This example was for illustrative purposes only and any connection to actual people, alive or dead, is strictly coincidental. :D

renata

14th August 2003, 02:01 PM

Kerberos

14th August 2003, 02:05 PM

Originally posted by Thanz

First, as is obvious (I think) I did not do any calculations to come up with my statement that 78 is inadequate. I used a common sense approach, and figured that with 26 possible guesses, 78 is not enough to get an accurate distribution across the whole set, which is what I perceive as "meaningful".
But there aren't 26 possible guesses - there are 2 rare and common letters

Originally posted by Thanz

As to the substance, I don't think that it is adequate to simply look at the mentions of one letter. I think what is required is a distribution of guesses across all possible outcomes (letters) and a comparison to the frequency in the normal population. And I still think that 78 is too small for such an analysis.
One letter? No one has analyzed the frequency of one letter, and in any case one letter could be enough if the results were statistically significant.

Originally posted by Thanz

Further, we need to look at the source of the 78 and see if we are comfortable with it as representative of the guesses he makes. We only have data from LKL, and I think one CO. CO data is problematic as it will probably over emphasize hits. LKL data is problematic as it would seem that that is the only place where he does these more rapid readings. If we had some data from seminars, for example, I think it would be much better. I am not convinced that 70-odd guesses from LKL are representative of JE readings in general.
I don't see why the LKL readings are of less value, since it really doesn't matter if he gets contact with the correct spirits. What matters is that he according to himself at least gets contact with some sort of spirits that gives him names or initials. If he truly gets into contact with spirits the initials of these names should be representative of the population as a group.

Originally posted by Thanz

One other point - I don't think that a guess of "John" can be equated with a guess of "J". The probability of getting a "J" is higher than getting a "John", which makes them different guesses. If "John" is a miss, and he goes on to "some other J name" then I would agree. I don't think that this is changing the statistics - I think it is making sure we get the right statistics in the first place.
Sure a John is a more impressive hit than a J, but hits never enter into the calculation. It might have confused you that I used hits to describe a guess of a rare letter, but it's totally irrelevant if the sitter validates the guess. The only thing that matters is the initial letter of the guess JE makes.

neofight

14th August 2003, 02:08 PM

Originally posted by jim_scotti

So I suppose it's just "mere speculation" when someone is critical of your hero, yet when a sitter on CO credulously supports JE, it is solid fact? While O'Neill might not have had proof, he sure had solid observations of what was going on and it doesn't take much circumstantial evidence to show that hot reading is very possible under those conditions.

Solid observations? Jim, how exactly would one go about solidly observing a hidden microphone? Come on now. Don't just jump to refute me merely because I'm a believer, and you are not. Use that critical thinking I know you possess. ;)

Well thanks, I see you are good at supplying references for me.

Jim.

Well, the sad truth of it is, Jim, that I am pathetically bad at doing searches, and I have a few of my own to do today. :( Otherwise, I would be very happy to help out a newbie, so please don't take it personally. :D

However, you are indeed in luck, because there are many posters here who excel at this, including CFLarsen and Instig8R, and maybe they'll be able to help you out.......neo

Kerberos

14th August 2003, 02:12 PM

Originally posted by renata

We shall just write this one off to dumb humorless Danes, OK? ;)

Hey! sterotyping isn't nice :hit:

neofight

14th August 2003, 02:13 PM

Originally posted by renata

Hey, Neo, Instig8r had some questions for you in the Appeal to Honest Skeptics thread about this very issue!! :)

Thank you for your personal interest, renata. :rolleyes: I believe I've already responded to everything that I intended to, but I'll give it a look-see, okay? :p .......neo

renata

14th August 2003, 02:16 PM

Originally posted by Kerberos

Hey! sterotyping isn't nice :hit:

Oh Lord! Another one!! :D

( I have nothing but respect and admiration for the dumb humorless Danes) ;)

CFLarsen

14th August 2003, 02:17 PM

neo,

Feel free to post those objectionable posts of Instig8r's, as well as the criticism.

And...stop asking other people to do your searches for you. You know damn well how to search, you just "choose" not to, especially if the results will show you wrong.

Clancie

14th August 2003, 02:49 PM

Posted by CFLarsen

Feel free to post those objectionable posts of Instig8r's, as well as the criticism.
Claus, do you really have to stick your nose in and try to fan the flames everytime someone here has a disagreement? I'm sure people are able to work this out for themselves without your usual goading...baiting...etc. Why not just MYOB instead of trying to keep stirring negative feelings up whenever you can?

CFLarsen

14th August 2003, 03:02 PM

Originally posted by Clancie

Claus, do you really have to stick your nose in and try to fan the flames everytime someone here has a disagreement? I'm sure people are able to work this out for themselves without your usual goading...baiting...etc. Why not just MYOB instead of trying to keep stirring negative feelings up whenever you can? [/B]

It's about answers, Clancie.

Got any?

Lurker

14th August 2003, 03:04 PM

Loki

14th August 2003, 04:38 PM

Neofight,

(while I leave kerberos, Bill and lurker to work out the statistics of the "rare letters" matter...)
Jim, how exactly would one go about solidly observing a hidden microphone?
One thing I've never quite understood is why JE supporters would ignore that fact that the CO Gallery is 100% covered by electronic monitoring equipment (audio and video) for the entire time the (potential) sitters are there. As part of the preparation for a taping session, all this equipment would be given a "dry run" to ensure it works. No "hidden" mircophones are required - the place is covered in clearly visible microphones.

How hard would it be to split a feed off to just *one* key staffer who monitors the general conversation in the Gallery before tapoing begins, and passes on a few bits of info to JE along with a general (or even specific) area. A good cold reader armed with just a few bits of "impossible to get from the public record" data, and backed by good post editing is going to look *very* good, isn't he? And this doesn't require the "entire staff" to be aware of this at all.

And I know you've offered the "but sitters will be suspicious if JE just repeats back to them things they said 5 minutes earlier" - but you're being too literal, aren't you? Take the "Niagara Falls" example - what if the woman had said to her companions "I love Buffalo this time of year..." as part of a general conversation while waiting. JE comes out, targets her, and asks "whats the Niagara Falls connection". Sure, he's taking a *bit* of a punt, but it's a reasonably safe bet! If JE is clever (I think he is) and a fake (I think he is) then he's certainly smart enough to be able to manufacture information without needed to just repeat what was said.

neofight

14th August 2003, 05:18 PM

Originally posted by Loki
Neofight,
One thing I've never quite understood is why JE supporters would ignore that fact that the CO Gallery is 100% covered by electronic monitoring equipment (audio and video) for the entire time the (potential) sitters are there. As part of the preparation for a taping session, all this equipment would be given a "dry run" to ensure it works. No "hidden" mircophones are required - the place is covered in clearly visible microphones.

Of course it is, Loki. It's a television studio. I agree with you. The point that I was trying to make is that it was all simply speculation on O'Neill's part, and for Jim to agree that O'Neill had "solid observations" of what was going on, when there wasn't really anything to observe, doesn't seem too objective. After all.......

from Michael O'Neill
"I think that the whole place is bugged somehow."

.....isn't exactly a specific or credible "observation", in my view.

Neither is assuming that simply because fifteen people were shuttled over to the studio in the same van, it meant that they were ringers for JE for heaven's sake. This is an especially flawed and inane piece of deductive reasoning, all the more so when one considers that O'Neill, of all people, was the one who was read. ;)

And I know you've offered the "but sitters will be suspicious if JE just repeats back to them things they said 5 minutes earlier" - but you're being too literal, aren't you? Take the "Niagara Falls" example - what if the woman had said to her companions "I love Buffalo this time of year..." as part of a general conversation while waiting. JE comes out, targets her, and asks "whats the Niagara Falls connection". Sure, he's taking a *bit* of a punt, but it's a reasonably safe bet! If JE is clever (I think he is) and a fake (I think he is) then he's certainly smart enough to be able to manufacture information without needed to just repeat what was said.

Is that possible? Of course it is. Did it happen? Why don't we ask Michael O'Neill if, while waiting for the show to begin, he spoke to anyone at all about anything that subsequently came through in the reading that JE did for him, or anything that JE might have used to make an educated guess on one of his hits. Oh, I forgot. We can't. Nobody ever heard from the mysterious Mr. O'Neill again, so we can't really follow up on any of these questions. :confused: ......neo

Kerberos

14th August 2003, 11:27 PM

Originally posted by renata

Oh Lord! Another one!! :D

( I have nothing but respect and admiration for the dumb humorless Danes) ;)
Gee thanks! :i:

Kerberos

14th August 2003, 11:27 PM

Originally posted by Lurker

At the risk of opening up a can of worms, I am totally with Thanz on this one.

For kicks and giggles I used the data from the census and compared to a sample size of 231 (where I work) and compared probabilities. Here are the differences I got:

Letter Difference (sample vs population):

A 60% (0.026 vs 0.065)
B 16% (0.039 vs 0.046)
C 56%
D 23%
E 53%
F 1%
G 7%
H 62% and so on.

While this analysis is anecdotal I think it clearly shows that using only a sample size of 78 people to define a population's first initial to their first name is woefully inadequate. At least if you want to have a reasonable precision to your numbers.

You're committing the same error as before, by analyzing each letter separately rather than as common and rare letter.

BillHoyt

15th August 2003, 04:34 AM

Originally posted by Lurker

At the risk of opening up a can of worms, I am totally with Thanz on this one.

For kicks and giggles I used the data from the census and compared to a sample size of 231 (where I work) and compared probabilities. Here are the differences I got:

Letter Difference (sample vs population):

A 60% (0.026 vs 0.065)
B 16% (0.039 vs 0.046)
C 56%
D 23%
E 53%
F 1%
G 7%
H 62% and so on.

While this analysis is anecdotal I think it clearly shows that using only a sample size of 78 people to define a population's first initial to their first name is woefully inadequate. At least if you want to have a reasonable precision to your numbers.

Lurker

Holy mackerel, Lurker, do you really wish to maintain that this analysis is correct in any sense? Do you know what you just did?

BillHoyt

15th August 2003, 05:52 AM

Originally posted by BillHoyt

Holy mackerel, Lurker, do you really wish to maintain that this analysis is correct in any sense? Do you know what you just did?

It occurs to me I probably lost 75% of the readership with this one, so I'll explain the comment a bit more for everybody, including Lurker.

Lurker, this is both mathematically and statistically wrong -- dead wrong. When I say "mathematically wrong", I mean something as basic as arithmetic. The way to see your error for yourself is to break down your calculation completely, into all of its components. The glaring nature of the arithmetic error should jump out at you. The "statistically wrong" part will become a moot point when you realize the mathematical error.

Cheers,

BillHoyt

15th August 2003, 04:09 PM

I guess Lurker is away. Does anybody else have any thoughts on where his analysis went wrong?

Cheers,

Instig8R

16th August 2003, 09:41 AM

Originally posted by neofight
quote:
--------------------------------------------------------------------------------
Originally posted by renata
LOL! Claus, silly, the references are in private messages! You would not want confidences to be broken, would ya?
--------------------------------------------------------------------------------

Boy, what a couple of butt-in-skis you two are. :rolleyes:

Claus, you, sir, are a liar, as per usual. Nothing new here. You excell at misrepresenting what others say, and by now, others know it too, believers and skeptics alike.

I never referred to "posts" on TVTalkShows in the context that you say I did. When I mentioned the "posts" it was in reference to the hateful language that Instig8R habitually used when discussing JE.

-snip-

Anything written to me privately, either by PM or e-mail, I am under no obligation to quote, and neither do I need it as evidence, since there were three individuals, including myself, who had already made public comments concerning Instig8R's nasty JE references.

So for once, Claus, why don't you try to get something right, as long as you are going to butt into other people's business anyhow?.....neo

Hi, Neo-- You have personally made some serious allegations. This is JREF, not "Pam's Board of Adoration of JE". In this forum, posters are expected to back-up their claims. You have not yet done so. Obviously, I am not the only person here who is interested in viewing the documentation of your claims.

Claus did not lie, and there is no valid basis for you to say that he did. Likewise, no one appears to be buying into your elaborate excuses to explain the issue of private messages, for the following reason: The issue of private messages (PM or e-mail) was brought up by YOU. There is a written record here, which verifies the chronology.

With respect to your most recent claim that you do not need private messages as evidence, here is what you really need to explain:

If you don't need private messages as evidence, WHY did YOU raise it as an issue???

To refresh your recollection, here is the dialogue:

Originally posted by Instig8R
P.S. I'd also like to know the identities of the other posters that you claim have commented on my "sometimes malicious language". Just who are these nameless posters?

Originally posted by neofight
As though knowing me, you really believe that I would state publicly what someone has told me privately. :rolleyes:
Please enlighten us as to why you felt the need to bring up the content of your private messages. Do you consider these anonymous character assassinations to be valid evidence of your claim? If so, I guess I can say that the weight of authority supports my position, because I have thus far received 27 private messages from other posters, confirming that you are being dishonest and are acting like a jerk.

Uhhh, what do you mean you want to know the identity of the posters and what they said? Hmmmph. As though knowing me, you reallyl believe I would state publicly what someone has told me privately. :rolleyes:

BillHoyt

16th August 2003, 09:24 PM

I re-worked the transcripts and came up with different results and a different method. Here it is, in a nutshell:

1. I used the census data figures orginally presented, although these may need tweaking.
2. I excluded the CO show data, and concentrated solely on the available, unedited transcripts from LKL, etc.
3. I looked at JE's style and adjusted the counting procedure as follows:
o I counted all of his name guesses
o Whether he stated them as names or initials, I counted them
o I excluded impossible-to-deal-with things such as "a B softened by a vowel," and chalked that up to a "B" guess.
o I included even bizarre names such as "pepper", "salt", "brooklyn" and other nickname guesses, except that
o I only counted "Liz", "Elizabeth" type guesses as the full given name, and did not also count an "L". but
o When JE recited a littany of names, I counted each one, whether they had the same initial or differing initials (again excluding the "Liz/Elizabeth, Ronny/Ronald, and Bill/William" type guesses, where I only counted the intial of the full given name.

Sound a bit complicated? You should read the transcripts. I could not see another way to approach things fairly given that sometimes he was all over the board. My hypothesis was, that, if there is a JE mediumship process, i should honor as much of it as I could figure in making the counting rules.

The results were a revised total guess count of 85. I then tallied the "J"s separately. I picked the "J"s because they are the most frequent initial. According to the US census data presented earlier, "J" surnames are 13.36% of the total population. In this analysis of 85 JE name guesses, I counted 18 "J" names. I calculated the expected number of "J"s (formally, the "expectation function") as 11.05.

I used the Poisson function to model the population. With an expected mean of 11.05, a count as high as (or higher than) 18 is expected to happen around 3% of the time. (Now remember we're rejecting the null hypothesis at 5%.) Based on that, I would reject the null hypothesis and say that this analysis refutes the hypothesis that JE's guesses are indistinguishable from purely random.

I'm double-checking my counts and calculations.

Cheers,

T'ai Chi

17th August 2003, 01:20 AM

Some thoughts:

After reading through all of this thread, I'm still not sure *exactly* what the question of interest is here.

Would this be your guys' null and alternative hypotheses:

Ho: JE says letters that match the expected letter frequencies
Ha: JE's says letters that are higher than or lower than the expected letter frequencies

,or something else? If so...

I'd suggest possibly looking into a chi-square goodness of fit analysis. Our response variable has > 2 categories (26 categories). We need to make sure each expected "cell count" (the number of readings*the proability for each letter) is greater than 5, for the analysis to be "good".

Our Ho and Ha would be as stated above. Written in more detail,

Ho: the proportion of A's JE says = the proportion of A's in the Census Bureau list
,the proportion of B's JE says = the proportion of B's in the Census Bureau list
,the proportion of C's JE says = the proportion of C's in the Census Bureau list
, ...
,the proportion of Z's JE says = the proportion of Z's in the Census Bureau list

Our test statistic is SUM[(Oi-Ei)^2/Ei], where Oi is the JE's frequency for the ith letter, and Ei is the Census Bureau frequency for the ith letter, where i runs from 1 to 26, representing A through Z.

We'd compare our test statistic to a chi-square with 25 = 26-1 degrees of freedom.

Doing this, we have statistically significant results (that is, we reject Ho, and conclude that JE's says letters that are higher than or lower than the expected letter frequencies) when our test statistic is larger than 37.7.

Late at night, I'm not sure if the Poisson analysis is correct. But something seems not right about it intuitively anyway because we should be interested in all the letters JE says, not just the J.

BillHoyt

17th August 2003, 06:50 AM

Originally posted by T'ai Chi
Some thoughts:

After reading through all of this thread, I'm still not sure *exactly* what the question of interest is here.

Would this be your guys' null and alternative hypotheses:

Ho: JE says letters that match the expected letter frequencies
Ha: JE's says letters that are higher than or lower than the expected letter frequencies

,or something else? If so...

I'd suggest possibly looking into a chi-square goodness of fit analysis. Our response variable has > 2 categories (26 categories). We need to make sure each expected "cell count" (the number of readings*the proability for each letter) is greater than 5, for the analysis to be "good".

Our Ho and Ha would be as stated above. Written in more detail,

Ho: the proportion of A's JE says = the proportion of A's in the Census Bureau list
,the proportion of B's JE says = the proportion of B's in the Census Bureau list
,the proportion of C's JE says = the proportion of C's in the Census Bureau list
, ...
,the proportion of Z's JE says = the proportion of Z's in the Census Bureau list

Our test statistic is SUM[(Oi-Ei)^2/Ei], where Oi is the JE's frequency for the ith letter, and Ei is the Census Bureau frequency for the ith letter, where i runs from 1 to 26, representing A through Z.

We'd compare our test statistic to a chi-square with 25 = 26-1 degrees of freedom.

Doing this, we have statistically significant results (that is, we reject Ho, and conclude that JE's says letters that are higher than or lower than the expected letter frequencies) when our test statistic is larger than 37.7.

Late at night, I'm not sure if the Poisson analysis is correct. But something seems not right about it intuitively anyway because we should be interested in all the letters JE says, not just the J.

The way you frame the hypothesis doesn't get to the issue. This would simply distinguish between random and non-random letters. We're specifically interested in the idea that the letter frequencies might give away cold-reading. Therefore, we're looking for higher hits on the more frequent letters or lower hits on the lower frequency letters.

Chi square won't work with frequencies less than 5. That makes it useless for this case.

Cheers,

Jeff Corey

17th August 2003, 07:04 AM

Originally posted by BillHoyt
Chi square won't work with frequencies less than 5. That makes it useless for this case.
Fisher's Exact Probability Test will.

BillHoyt

17th August 2003, 07:53 AM

Originally posted by Jeff Corey

Fisher's Exact Probability Test will.

Arggh! A 26 x 26 matrix? Arggh!

I think this suffers from the problem of looking at the sameness of the distributions. What I would be after here is to know if he focuses on the frequent initials and ignores the less frequent.

Cheers,

T'ai Chi

17th August 2003, 09:48 AM

Originally posted by BillHoyt

The way you frame the hypothesis doesn't get to the issue. This would simply distinguish between random and non-random letters. We're specifically interested in the idea that the letter frequencies might give away cold-reading. Therefore, we're looking for higher hits on the more frequent letters or lower hits on the lower frequency letters.

Chi square won't work with frequencies less than 5. That makes it useless for this case.

Cheers,

I framed the hypothesis the way I did Bill, because I still didn't know exactly what the hypothesis that people on this thread had.

I'd probably decide just what differentiates between "high" and "low" frequency letter. Decide what this cutoff percentage is. Doing this objectively could be difficult, and it is always possible that the final analysis results could be sensitive to this choice. Something to think about.

Then I would perhaps think about doing two chi-square goodness of fit tests, one GOF with just the high frequency letters, and one GOF with just the low frequency letters. I'd initially only do the high frequency letters, and see how that goes.

A chi-square goodness of fit test will certainly work if some expected cell counts are less than 5, it just won't provide a totally reliable p-value because the approximation isn't as good as it could be. In any case, starting out with just the high frequency letters could avoid this issue.

BillHoyt

17th August 2003, 09:56 AM

Originally posted by T'ai Chi

I framed the hypothesis the way I did Bill, because I still didn't know exactly what the hypothesis that people on this thread had.

I'd probably decide just what differentiates between "high" and "low" frequency letter. Decide what this cutoff percentage is. Doing this objectively could be difficult, and it is always possible that the final analysis results could be sensitive to this choice. Something to think about.

Then I would perhaps think about doing two chi-square goodness of fit tests, one GOF with just the high frequency letters, and one GOF with just the low frequency letters. I'd initially only do the high frequency letters, and see how that goes.

A chi-square goodness of fit test will certainly work if some expected cell counts are less than 5, it just won't provide a totally reliable p-value because the approximation isn't as good as it could be. In any case, starting out with just the high frequency letters could avoid this issue.

T'ai,

The point is that chi won't work on the low frequencies. The p value is just a bit important. We have frequencies in the control population of .12, .01, .022 and we have a sample population of around 80. Ain't gonna happen.

Cheers,

T'ai Chi

17th August 2003, 10:11 AM

Originally posted by BillHoyt

T'ai,

The point is that chi won't work on the low frequencies. The p value is just a bit important. We have frequencies in the control population of .12, .01, .022 and we have a sample population of around 80. Ain't gonna happen.

Cheers,

What are the high frequency letters, and what are the frequencies for the high-frequency letters?

BillHoyt

17th August 2003, 10:40 AM

Originally posted by T'ai Chi

What are the high frequency letters, and what are the frequencies for the high-frequency letters?

See this post. (http://host.randi.org/vbulletin/showthread.php?s=&postid=1870048960&highlight=census#post1870048960)

Cheers,

T'ai Chi

17th August 2003, 11:54 AM

Originally posted by BillHoyt

See this post. (http://host.randi.org/vbulletin/showthread.php?s=&postid=1870048960&highlight=census#post1870048960)

Cheers,

Thanks, I understand that Bill. I am also asking what distinguishes a "high frequency letter"? What is the cutoff that differentiates a high frequency letter from a low frequency letter, and can it be done objectively, and how sensitive is the final analysis to this cutoff?

The post you link to above, what are those numbers? Many are greater than 1. Are these percentages, so 7.5, for example, would be 7.5%, or .075?

17th August 2003, 12:04 PM

Originally posted by T'ai Chi

Thanks, I understand that Bill. I am also asking what distinguishes a "high frequency letter"? What is the cutoff that differentiates a high frequency letter from a low frequency letter, and can it be done objectively, and how sensitive is the final analysis to this cutoff?

The post you link to above, what are those numbers? Many are greater than 1. Are these percentages, so 7.5, for example, would be 7.5%, or .075?

I suggest you read the article again, and learn from it for yourself, instead of insisting that others who you refuse to listen to actually manage to teach you.

T'ai Chi

17th August 2003, 12:36 PM

Ignoring jj's non-help. I'm actually interested in statisically analysing JE's letter use.

Has anyone revealed what distinguishes a high frequency letter from a low frequency letter?

Anyway, how many instances of JE choosing letters do we have again, 78?

I'll use the letters where 78*p > 5, as my criteria for high frequency until someone has a more sensible one, and I'm sure a better criteria exists.

letter, expected frequency
a, .064861
c, .072108
d, .074201
j, .133615
m, .100377
r, .08003
(s was close to having 78*p around 5, and l's 78*p was over 4)

If we had a larger n, for example, n = 100, we'd have 8 high frequency letters instead of 6.

So just looking at the 6 high frequency letters, I would do a chi-square goodness of fit test, and compare the test statistic to a chi-square distribution on 5 = 6-1 degrees of freedom. For the results to be statistically significant at the 5% level, we'd need our test statistic to be larger than 11.07.

So does anyone know the actual frequency of JE's usage of a, c, d, j, m, and r?

neofight

17th August 2003, 05:39 PM

Originally posted by Instig8R

Hi, Neo-- You have personally made some serious allegations. This is JREF, not "Pam's Board of Adoration of JE". In this forum, posters are expected to back-up their claims. You have not yet done so. Obviously, I am not the only person here who is interested in viewing the documentation of your claims.

Hi, Instig8R. I'll respond here to this last post of yours, but since you thought it necessary to move this issue over to my thread, "An Appeal To Honest Skeptics", you can find all my other responses over there. Unless, that is, you intend to introduce this subject into multiple threads in an effort to make some kind of a point or something. :rolleyes:

First of all, what were these "serious" allegations that I made against you? I stated that you made some malicious (Synonyms: bitchy, catty, despiteful, evil, hateful, malevolent, malign, malignant, nasty, rancorous, spiteful, spitish, vicious, wicked) comments about JE, and that I didn't care for them. Bottom line is that I don't have to care for them. Many people here and even over at TVTalkShows make remarks that are just as offensive, perhaps more so. That is not the point.

The point is that if you make such comments, other posters have a perfect right to give you their own opinion on what they think about you making them, being that you make them without any sort of proof that they are true. That's all I've ever done, argued with you about the rightness or wrongness of making such derogatory comments when you do not have any proof that there is any truth to them whatsoever.

Other than that, what else did I allege? That you have an anti-JE bias? lol Oh, and you're here to say that you don't? Please. I've already addressed this over in the other thread. Yes, yes, I know. Everybody here is arguing from the standpoint of one bias or another. I know. So why did you dispute the fact, and say that I was making it all up? You say you don't have an excessive anti-JE bias? Well, sorry, 'g8R, but that is debatable. I feel that you do.

Claus did not lie, and there is no valid basis for you to say that he did.

You know, I'm inclined to agree with you about this particular statement, and say that I think perhaps Claus did not intentionally lie this time. Claus did, I'm afraid, get the facts wrong, as per usual, but after going back and re-reading one of your own posts, I see that he was just careless, and he relied on your word for something instead of going back and reading my own original post. Unfortunately, you didn't represent my words accurately, so while it may have been careless of Claus to get his information from you, rather than me, I don't believe it was intentional on his part, and I am very pleased to be able to say that. :)

Likewise, no one appears to be buying into your elaborate excuses to explain the issue of private messages, for the following reason: The issue of private messages (PM or e-mail) was brought up by YOU. There is a written record here, which verifies the chronology.

Actually, Instig8R, it looks as though you might be attempting to do a little manipulating of your own here. What actually happened is that I made a reference to other posters who felt as I did, and I was mainly thinking of atmytv and his apology when I said that. He and I seem to feel pretty much the same way on this issue, as you can see from his following quote.......

(atmytv)"In fact, I understand how people who think John is not real would feel that he is defrauding people and is scum. However, for those people to state it as if it is a fact and throw it at us is insensitive at best."

I hadn't mentioned any names, however, but in your next post, you, yourself, brought up atmytv and sgrenard's names, sort of like a preemptive strike I imagine, or because you seem to have the opinion that simply because atmytv subsequently made an apology to you, (mainly because you appeared to have been personally offended by what he had said) you thought perhaps that he had recanted, when, indeed, he had done no such thing. He merely said that in the future, he would try to be more good-natured about the matter.

With respect to your most recent claim that you do not need private messages as evidence, here is what you really need to explain:

If you don't need private messages as evidence, WHY did YOU raise it as an issue???

Again, the only reason that they came up at all, was because you had already acknowledged atmytv and sgrenard, and were demanding what other posters I was referring to. Well, frankly, they were the posters that I was going to cite, since they were the ones who had criticised you publicly. I never intended to even mention anyone else. But........

you insisted on my giving you the other names and quotes, and I said that I would not, because they had not chosen to make their opinions public. YOU pushed it. Not me. :confused:

Do you consider these anonymous character assassinations to be valid evidence of your claim? If so, I guess I can say that the weight of authority supports my position, because I have thus far received 27 private messages from other posters, confirming that you are being dishonest and are acting like a jerk.

Uhhh, what do you mean you want to know the identity of the posters and what they said? Hmmmph. As though knowing me, you reallyl believe I would state publicly what someone has told me privately. :rolleyes:

Instig8R, consider where we are posting. I take it for granted that some people here may think that, and I have no interest whatsoever in knowing who they might be, mainly because a good number of them have already made their opinions of me, and believers in general, pretty darned clear. :rolleyes: BFD

Now, if you don't mind, if you have anything further to add, please post it over on the "AATHS" thread. Thanks! :) .......neo

Lurker

18th August 2003, 04:08 AM

Originally posted by Kerberos

You're committing the same error as before, by analyzing each letter separately rather than as common and rare letter.

Actually, I knew that. I was just curious how well 78 would hold up in the problem that I had originally thought was being proposed. As I stated previosuly, 78 may be enough for 2 possible outcomes but clearly not enough for 26 possible outcomes.

Lurker

Lurker

18th August 2003, 04:11 AM

Originally posted by BillHoyt

Holy mackerel, Lurker, do you really wish to maintain that this analysis is correct in any sense? Do you know what you just did?

What? Oh no! I don't KNOW what I just did?

Lurker

BillHoyt

18th August 2003, 04:56 AM

Originally posted by Lurker

What? Oh no! I don't KNOW what I just did?

Lurker

Apparently not, Lurker.

Kerberos

18th August 2003, 05:31 AM

Originally posted by Lurker

What? Oh no! I don't KNOW what I just did?

Lurker

Originally posted by BillHoyt

Apparently not, Lurker.

Can't you just tell what he did wrong instead of playing guessing games?

Lurker

18th August 2003, 06:01 AM

Yes, I am all ears. Always willing to learn more...

Bill? Done playing coy?

Lurker

Thanz

18th August 2003, 06:40 AM

Originally posted by BillHoyt

T'ai,

The point is that chi won't work on the low frequencies. The p value is just a bit important. We have frequencies in the control population of .12, .01, .022 and we have a sample population of around 80. Ain't gonna happen.

Cheers,

Hmmm.....

Perhaps the problem is small sample size??

CFLarsen

18th August 2003, 06:48 AM

Originally posted by Thanz
Hmmm.....

Perhaps the problem is small sample size??
Hmmm.....

Perhaps you could solve the problem by telling us how bit the sample size should be?

Dazzle us with your knowledge of statistics.

BillHoyt

18th August 2003, 06:50 AM

Originally posted by Thanz

Hmmm.....

Perhaps the problem is small sample size??

I invite you to read the posts carefully. Note that T'ai is proposing a particular statistical test, one of many possible. Note that I commented solely on the one-tailed chi-squared test for the low frequency initials.

Cheers,

BillHoyt

18th August 2003, 06:56 AM

Originally posted by Lurker
Yes, I am all ears. Always willing to learn more...

Bill? Done playing coy?

Lurker

I was hoping the realization would come to you, Lurker. Let me pose this question to shed some light on the problem:

If we can take .026 of population A and .065 of population B and declare them 60% different, then why does taking .026 of an inch and .065 of a foot and declaring them 60% different immediately trigger the error bell?

Cheers,

Lurker

18th August 2003, 07:09 AM

Originally posted by BillHoyt

I was hoping the realization would come to you, Lurker. Let me pose this question to shed some light on the problem:

If we can take .026 of population A and .065 of population B and declare them 60% different, then why does taking .026 of an inch and .065 of a foot and declaring them 60% different immediately trigger the error bell?

Cheers,

Sorry, I am a bit slow. But I consider your comparison worthless. I am examining rates. Please provide a better analogy or just speak plainly so those of us who do not have your prodigious intellect can understand you. Thank you.

Keep in mind my population B is a subset of Population A. A Sample if you will. If you want to poke holes in being unrepresentative I will agree with you. Clearly a workplace will have more of certain age groups and not represent others. Other than age, race and economic factors may skew the sample size.

Let me make an analogy so you can understand rates. Let us say the US murder rate is 5% (quantified over some period of time). In Houston let us say the muder rate is 7%. By using math we can see that (7-5)/5 = 0.4. Thus, we can say that the murder rate in Houston is 40% higher than the murder rate in the US. Note we are not adding 40% to the murder rate but are making a comparison? Capiche?

Lurker

Lurker

BillHoyt

18th August 2003, 07:14 AM

Originally posted by Lurker

Sorry, I am a bit slow. But I consider your comparison worthless. I am examining rates. Please provide a better analogy or just speak plainly so those of us who do not have your prodigious intellect can understand you. Thank you.

Keep in mind my population B is a subset of Population A. A Sample if you will. If you want to poke holes in being unrepresentative I will agree with you. Clearly a workplace will have more of certain age groups and not represent others. Other than age, race and economic factors may skew the sample size.

Let me make an analogy so you can understand rates. Let us say the US murder rate is 5% (quantified over some period of time). In Houston let us say the muder rate is 7%. By using math we can see that (7-5)/5 = 0.4. Thus, we can say that the murder rate in Houston is 40% higher than the murder rate in the US. Note we are not adding 40% to the murder rate but are making a comparison? Capiche?

Lurker

Lurker

Fine. Let's turn my example into rates. (And after reading this, please calm down before you post again. I set you up and you bit without even thinking. Cool down a bit before your next post) Does comparing .026 inches per hour and .065 feet per hour and declaring them 60% different make any sense to you?

Capiche?

Thanz

18th August 2003, 07:15 AM

Originally posted by BillHoyt

I invite you to read the posts carefully. Note that T'ai is proposing a particular statistical test, one of many possible. Note that I commented solely on the one-tailed chi-squared test for the low frequency initials.

Cheers,
The reason I posted is because the only objection that you had to T'ai Chi's proposed test was that it would not work with the sample size we have available. That does not tell me that T'ai Chi's test itself isn't appropriate for the work we want to do - it just tells me that we can't do it with the sample size we have now.

If I am reading Tai Chi's posts correctly, it would seem that to do his analysis one needs to have a sufficient sample size so that you would expect each letter to appear at least five times. Is that correct?

BillHoyt

18th August 2003, 07:17 AM

Now, then: what is a "rate"? What is a "percentage"? Break them down into their component parts. Why can you not do what you did at the fundamental level?

Cheers,

BillHoyt

18th August 2003, 07:19 AM

Originally posted by Thanz

The reason I posted is because the only objection that you had to T'ai Chi's proposed test was that it would not work with the sample size we have available. That does not tell me that T'ai Chi's test itself isn't appropriate for the work we want to do - it just tells me that we can't do it with the sample size we have now.

If I am reading Tai Chi's posts correctly, it would seem that to do his analysis one needs to have a sufficient sample size so that you would expect each letter to appear at least five times. Is that correct?

That's the analysis he proposed for the hypothesis he proposed. That's neither my analysis nor my hypothesis. I have already demonstrated that not only can one use Poisson to test JE's "J" frequency; the result is statistically significant and (significantly) refutes my null hypothesis.

Cheers,

Lurker

18th August 2003, 07:22 AM

Originally posted by BillHoyt
Now, then: what is a "rate"? What is a "percentage"? Break them down into their component parts. Why can you not do what you did at the fundamental level?

Cheers,

Sorry, still not getting through to me. I must be very dimwitted indeed. In the context that I am using the terms, they are interchangeable.

When I see 0.046 for the letter E from the census I interpret that to mean 4.6% of the people have a name starting with E. Or a rate , 4.6 people out of a hundred would have E names if I started counting.

Sorry, Bill. I guess I am a very poor student.

Lurker

BillHoyt

18th August 2003, 07:30 AM

Originally posted by Lurker

Sorry, still not getting through to me. I must be very dimwitted indeed. In the context that I am using the terms, they are interchangeable.

When I see 0.046 for the letter E from the census I interpret that to mean 4.6% of the people have a name starting with E. Or a rate , 4.6 people out of a hundred would have E names if I started counting.

Sorry, Bill. I guess I am a very poor student.

Lurker

Lurker,

What is a percentage? What went wrong with saying .026 inches per hour is 60% different from .065 feet per hour?

Cheers,

Kerberos

18th August 2003, 07:43 AM

Lurker

18th August 2003, 07:44 AM

Um.. your units differ? Inches versus feet? You did not divide by 12 to make the units the same. But as far as I can tell, my units are people for both calculations.

Good thing you are a bouncer in a bar. You would sure make a difficult teacher.

Lurker

Lurker

18th August 2003, 07:48 AM

Originally posted by Kerberos

If it's any conciliation I don't get it either.

Point noted. I don't understand why Bill is dragging this out so long. He could have been quick about it and posted his answers back on the 15th. Instead, it is the 18th and I STILL don't know what I did wrong.

Instead, some 69 hours have passed and about 6 posts later and I still have nothing but cryptic posts from Bill.

Patiently waiting for the enlightenment...

Lurker

BillHoyt

18th August 2003, 07:54 AM

Originally posted by Lurker
Um.. your units differ? Inches versus feet? You did not divide by 12 to make the units the same. But as far as I can tell, my units are people for both calculations.
Now you're onto it. The units are different. A percentage is, like a rate, actually a numerator divided by a denominator. In doing calculations with them, one must take care to keep in mind the units being carried around.

In doing the 60% calculation, you assumed one population is a subset of another. But that is circular reasoning; you assumed an answer to the question under test. In reality, in one case, your denominator for one is "population A" and the for the other is "population B".

Cheers,

Lurker

18th August 2003, 08:19 AM

Originally posted by BillHoyt

In doing the 60% calculation, you assumed one population is a subset of another. But that is circular reasoning; you assumed an answer to the question under test. In reality, in one case, your denominator for one is "population A" and the for the other is "population B".

Sorry Bill. I STILL don't get it. Why is it I constantly see other polls, even comparisons of polls. Clearly they are SAMPLES of a larger population. By your logic, there would be no such thing as comparing SAMPLES to the population at large.

You are going to have to be MUCH clearer in what you write.

Remember, population B is a subset of population A or a SAMPLE, if you will.

Lurker

Lurker

18th August 2003, 08:23 AM

Funny how I hear commentators say things along the line of "this is a 50% increase over the previous poll". Clearly two different polls are of two different samples, right? SAMPLE A and SAMPLE B. Or as you would term it, POPULATION A and POPULATION B.

I think you, Bill, and I are talking past each other again. I think I have plainly stated my case but am still having trouble understanding yours. Do you have a case to put forward here?

Lurker

Lurker

18th August 2003, 08:27 AM

Originally posted by BillHoyt

you assumed one population is a subset of another. But that is circular reasoning;

Um, the data provided came from the census bureau, which presumably provides the best and most complete data on names in the US.

I am comparing it to data on my company's personnel. Everyone is a US citizen here so unless they have not been counted by the census there certainly ARE a subset of the aforementioned data.

You can quibble about what they report on the census versus what their names are here but it seems pretty minor.

So, please explain how this is circular reasoning?

Gracias,

Lurker

BillHoyt

18th August 2003, 08:29 AM

Originally posted by Lurker

Sorry Bill. I STILL don't get it. Why is it I constantly see other polls, even comparisons of polls. Clearly they are SAMPLES of a larger population. By your logic, there would be no such thing as comparing SAMPLES to the population at large.

You are going to have to be MUCH clearer in what you write.

Remember, population B is a subset of population A or a SAMPLE, if you will.

Lurker

I'm tackling one layer at a time here. Population B is indeed a sample of A, and its representativeness is one of the base questions. That means I need to expand the "units" of the denominators. So let's try to translate those units: The census data remains population A, although we know it is not 100% accurate. This has been an issue for decades with the census and Congress has had a tizzy fit whenever the Census Bureau has suggested using statistical techniques to try to correct it. The implication here is that, even population A has a +/- to it. We know, of course, this is true of the sample of A that I had referred to as population B. This also has a +/- to it, and a larger one at that.

With that in mind, the denominators are actually "population A +/-X" and "population A +/-Y", where Y > X. We can do similar things with the standard deviation, kurtosis and skew. Any way we slice it, though, those denominators are not truly the same, and it is the sameness of those denominators that is the original question.

Cheers,

Lurker

18th August 2003, 08:34 AM

Originally posted by BillHoyt

you assumed one population is a subset of another. But that is circular reasoning; you assumed an answer to the question under test. In reality, in one case, your denominator for one is "population A" and the for the other is "population B".

Cheers,

So, provide your evidence that POP A is not a subest fo POP B. I have provided my logic for why it is. I humbly await your refutation of same.

Lurker

Lurker

18th August 2003, 08:40 AM

Originally posted by BillHoyt

I'm tackling one layer at a time here. Population B is indeed a sample of A, and its representativeness is one of the base questions. That means I need to expand the "units" of the denominators. So let's try to translate those units: The census data remains population A, although we know it is not 100% accurate. This has been an issue for decades with the census and Congress has had a tizzy fit whenever the Census Bureau has suggested using statistical techniques to try to correct it. The implication here is that, even population A has a +/- to it. We know, of course, this is true of the sample of A that I had referred to as population B. This also has a +/- to it, and a larger one at that.

With that in mind, the denominators are actually "population A +/-X" and "population A +/-Y", where Y > X. We can do similar things with the standard deviation, kurtosis and skew. Any way we slice it, though, those denominators are not truly the same, and it is the sameness of those denominators that is the original question.

I understand what you have written and to a certain degree I agree with it.

But, then no meaningful poll can ever be conducted. Your attempts to quantify the JE hits on first letters versus census data is impossible. So why are YOU advocating just such a comparison? IF the census numbers are hopelessly inaccurate and any sample (Like JE's 78) is also hopelessly inaccurate according to your definitions, what is the point? Why are you wasting time on it?

Lurker

B

BillHoyt

18th August 2003, 09:00 AM

Originally posted by Lurker

I understand what you have written and to a certain degree I agree with it.

But, then no meaningful poll can ever be conducted. Your attempts to quantify the JE hits on first letters versus census data is impossible. So why are YOU advocating just such a comparison? IF the census numbers are hopelessly inaccurate and any sample (Like JE's 78) is also hopelessly inaccurate according to your definitions, what is the point? Why are you wasting time on it?

Lurker

B

I'm not saying they are hopelessly inaccurate. Nor am I saying you can't compare data. I am saying you can't take the percentages of different things and compare them the way you did. (You can't take 50% of an apple, add it to 50% of an orange and get 100% of X. Unless X is a snack.) You can't simply look at 2.6% and 6.5% and say they are 60% different without knowing you are really comparing apples with apples. Percentages have units buried within them. When speaking of populations, they also have standard deviations, measurement errors, skew, and kurtosis built in. This is one of the reasons for statistics; to be able to deal with these animals given that those denominators are different.

Cheers,

Lurker

18th August 2003, 09:27 AM

Originally posted by BillHoyt

I'm not saying they are hopelessly inaccurate. Nor am I saying you can't compare data. I am saying you can't take the percentages of different things and compare them the way you did. (You can't take 50% of an apple, add it to 50% of an orange and get 100% of X. Unless X is a snack.) You can't simply look at 2.6% and 6.5% and say they are 60% different without knowing you are really comparing apples with apples. Percentages have units buried within them. When speaking of populations, they also have standard deviations, measurement errors, skew, and kurtosis built in. This is one of the reasons for statistics; to be able to deal with these animals given that those denominators are different.

Cheers,

Agreed again. But I was merely comparing averages. And clearly I was not comparing apples to oranges. I was comparing oranges from one shipment to all the known oranges. Methinks you are being unreasonably pedantic in your definitions.

I note you did not have the same problems when discussing the PASS/FAIL test for the rare/common letter test. Why do you think that is...

Lurker

Thanz

18th August 2003, 09:32 AM

I did a google search and found this nifty little Poisson calculator
http://hyperphysics.phy-astr.gsu.edu/hbase/math/poifcn.html
If I plug in BillHoyt's numbers, I get this:

If the probability of a single event is p =.1336 and there are n = 85 events, then the value of the Poisson distribution function at value x =18 is 1.802468785965208 x 10^ -2. For these conditions, the mean number of events is 11.356 and the standard deviation is 3.1366922705295783.

If I plug in Lurker's numbers for "A", I get this:

If the probability of a single event is p =.065 and there are n = 231 events, then the value of the Poisson distribution function at value x =6 is 4.796096105218159 x 10^ -3. For these conditions, the mean number of events is 15.015 and the standard deviation is 3.7468686926552417.

I would seem that the numbers for Lurker's office are even more statistically significant than for JE's J guesses. I think that this may be support for the argument being made by both Lurker and myself regarding small sample sizes and the meaningfullness of the analysis.

edited to fix link

BillHoyt

18th August 2003, 09:38 AM

Originally posted by Lurker

Agreed again. But I was merely comparing averages. And clearly I was not comparing apples to oranges. I was comparing oranges from one shipment to all the known oranges. Methinks you are being unreasonably pedantic in your definitions.

I note you did not have the same problems when discussing the PASS/FAIL test for the rare/common letter test. Why do you think that is...

Lurker

I'm hardly being pedantic here. You simply can't compare percentages unless you know the denominators are truly the same. If one could do this, then why would one bother with the hassle of inferential statistics? Why not simply set a percentage difference level and do what you did?

Actually, I did have the same problem with the rare/common letter test after all the specifics were discussed. The problem there, though, lies in the combined male/female data. I'm looking into the Census data to try to get those denominators so that the combined male/female percentages are done properly.

Cheers,

Lurker

18th August 2003, 10:03 AM

OK, Bill. Let's just go back and say I was a bit more precise and when I made the comparison I meant I was comparing the means. Does that help? I never meant to imply that my "for kicks and giggles" test was any sort of statistical analysis. All I did was comapre a SAMPLE versus the census. Nothing more, nothing less.

And if you want to make the claim that the confidence intervals created by the standard deviations are too high that is your option.

So, after all this talk do you have any interest in forming an opinion on how big a SAMPLE it would take to form a fairly accurate (you decide the parameters) histogram compared to census data? I still stand by it would take far more than 78 unless you are content with data that is all washed over.

Lurker

BillHoyt

18th August 2003, 10:18 AM

Originally posted by Thanz
I did a google search and found this nifty little Poisson calculator
http://hyperphysics.phy-astr.gsu.edu/hbase/math/poifcn.html (http://)

If I plug in BillHoyt's numbers, I get this:

If I plug in Lurker's numbers for "A", I get this:

I would seem that the numbers for Lurker's office are even more statistically significant than for JE's J guesses. I think that this may be support for the argument being made by both Lurker and myself regarding small sample sizes and the meaningfullness of the analysis.

Check your numbers again, Thanz. Also, check that calculator. For Poisson, the mean = the variance. The standard deviation is the square root of the variance. Those reported std devs. are off. More importantly, though, you reported the pdfs, not the cdfs of the tails of interest. Even more important than that, however, is that a test that confirms the null hypothesis does not mean the test procedure is wrong or that the sample size is wrong!

This is fundamental to understanding testing and statistics. If we go through the alphabet with a significance level of .05 and we test each letter individually, and the sample perfectly reflects the population, we will still expect to get a null hypothesis rejection.

Now let's do what you just did for "B", but get the tail and its cdf straight. Lurker reported .039, meaning he counted 9 "B"s. The expected percentage is .046, or 10.626 for this sample size. Of interest here is the left-tail, or the probability of this few observed or less. That is, .383. We accept the null hypothesis here, for B.

Cheers,

BillHoyt

18th August 2003, 10:32 AM

Originally posted by Lurker
OK, Bill. Let's just go back and say I was a bit more precise and when I made the comparison I meant I was comparing the means. Does that help? I never meant to imply that my "for kicks and giggles" test was any sort of statistical analysis. All I did was comapre a SAMPLE versus the census. Nothing more, nothing less.

And if you want to make the claim that the confidence intervals created by the standard deviations are too high that is your option.
I'm running out of ways to get the point across to you, Lurker. The "means"? How does that change the denominator problem? The denominator problem is arithmetic. You can't take 50% of an apple, add it to 50% of an orange and get 100% of anything.

Here is the Australian government making the same point to people handling their census data (emphasis mine):

"Any Indigenous statistical comparisons made between two censuses must be made with caution and should not be accepted at face value until the user has explored, to his/her satisfaction, the possibility that the differences might be solely or largely a consequence of non-demographic increase in census counts. Failure to do this could lead users to draw incorrect conclusions about whether changes in social conditions have occurred.

2 Use percentages

Users should present their statistical estimates as percentages where both numerator and denominator are data from the same census. Analyses of intercensal statistical differences should be made by comparing percentages from two times, rather than directly comparing counts or numbers. In most instances appropriate percentages will be less biased than the numerator and denominator counts. In particular, percentages are estimated without bias, if the bias in the counts is the same in percentage terms for the numerator and denominator."

source (http://www.abs.gov.au/websitedbs/c311215.nsf/0/eab052ee179fbe6bca256d8100201564?OpenDocument)
So, after all this talk do you have any interest in forming an opinion on how big a SAMPLE it would take to form a fairly accurate (you decide the parameters) histogram compared to census data? I still stand by it would take far more than 78 unless you are content with data that is all washed over.

Lurker
I have already commented on this. It depends on a large number of choices. Most particularly the choice of hypothesis.

Cheers,

Thanz

18th August 2003, 10:33 AM

Originally posted by BillHoyt
Check your numbers again, Thanz. Also, check that calculator. For Poisson, the mean = the variance. The standard deviation is the square root of the variance. Those reported std devs. are off. More importantly, though, you reported the pdfs, not the cdfs of the tails of interest. Even more important than that, however, is that a test that confirms the null hypothesis does not mean the test procedure is wrong or that the sample size is wrong!
I understand that, but my point is that small sample sizes will give strange, unreliable results.

What is the hypothesis here - for Lurker's office?

I say that there is some sort of "A-hole" located somewhere in Lurker's office. If someone with a name that starts with "A" gets too close, they are sucked into the "A-hole" never to be heard from again.

Therefore, I hypothesize that the number of people in Lurker's office with names that start with "A" will be under-represented when compared to the normal population.

Let's look at the data. Hmmm.... it looks like it supports my hypothesis quite strongly!

Look out Lurker! There is an "A-hole" somewhere in your office!!

Now let's do what you just did for "B", but get the tail and its cdf straight. Lurker reported .039, meaning he counted 9 "B"s. The expected percentage is .046, or 10.626 for this sample size. Of interest here is the left-tail, or the probability of this few observed or less. That is, .383. We accept the null hypothesis here, for B.
I don't know what the calculator is doing, but when I plug those numbers in I get a mean of 10.626, but a standard dev. that is not the square root of that number.

Lurker

18th August 2003, 10:45 AM

>You can't take 50% of an apple, add it to 50% of an orange and >get 100% of anything.

Why do you persist in such a silly comparison? I think most here can see that your comaprison is invalid.

>2 Use percentages

>Users should present their statistical estimates as percentages >where both numerator and denominator are data from the >same census. Analyses of intercensal statistical differences >should be made by comparing percentages from two times, >rather than directly comparing counts or numbers. In most >instances appropriate percentages will be less biased than the >numerator and denominator counts. In particular, percentages >are estimated without bias, if the bias in the counts is the same >in percentage terms for the numerator and denominator."

Need I point out that I DID use percentages? I think YOU are misinterpreting the Aussies here. I did not compare numbers, but percentages just like advised above. You are clearly confused if you thought otherwise. Why did you post #2 here? Did you not realize I was referring to percentages? Please explain cause I cannot see why you would post #2.

Further, I note you again avoided attempting to provide your opinion on how many it might take. If you have done so, please repeat the ballpark number. If not, why not? We' re not requesting a thesis here, just a freakin opinion.

Lurker