Analyzing Medium/Cold ReaderTranscripts

[T'ai Chi]

What do people think are some valuable things to analyze from the transcripts from self proclaimed mediums and from self proclaimed cold readers?

In this stage I'm building up a list of interesting things to study.

The next stage will be obtaining transcripts, but I need help. I'd like to get hundreds of transcripts ideally, from a variety of mediums and cold readers.

The stage after that will be describing the data (hopefully by assigning a few randomly chosen transcripts to randomly chosen volunteers for them to summarize numerically), using the key things from the first stage. Ideally building a model would be nice, but right now the model part is most definitely overly ambitious. This current project is entirely descriptive in nature.

Please PM me if you are interested in being a volunteer, and please mention whether you would like to help find transcripts, analyze transcripts, or both.
(although please don't start doing these things)

This might be a year-long project or longer, but I think it would be really interesting to undertake if there is enough interest.

[CFLarsen]

How should we analyze them objectively? Based on what criteria?

[Paul C. Anagnostopoulos]

How are you going to calculate the probability that a good hit was simply chance?

~~ Paul

[Ed]

Just try it yourself.

[Ersby]

Originally posted by T'ai Chi
I'd like to get hundreds of transcripts ideally, from a variety of mediums and cold readers.


That's not going to happen. JE is the foremost medium today, and I've only got about 30 transcripts of him. Asking for hundreds is too tall an order.

To answer your question, personally I think names would be a good area to examine. Not in a letter count type of way (let's not go down that road again) but because guesses about names are easily quantifiable.

[Clancie]

Posted by Ersby

To answer your question, personally I think names would be a good area to examine. Not in a letter count type of way (let's not go down that road again) but because guesses about names are easily quantifiable.
Yes, although no one seemed impressed by the various hits on unique names Piper got in the transcript excerpt I quoted from Gauld.

My guess is that (1) we won't find a cold reader who does very well with names and (2) if we find a medium who does well with names it will be discounted as sitter buy-in, lucky guessing on common names, and/or cheating.

[CFLarsen]

Originally posted by Clancie
Yes, although no one seemed impressed by the various hits on unique names Piper got in the transcript excerpt I quoted from Gauld.

My guess is that (1) we won't find a cold reader who does very well with names and (2) if we find a medium who does well with names it will be discounted as sitter buy-in, lucky guessing on common names, and/or cheating.

So, how do you suggest we investigate and compare transcripts from cold readers and psychic mediums, then?

Don't just criticize, come up with better suggestions.

[Clancie]

T'ai,

That sounds like a very ambitious project, especially given the lack of cold reading resource material (at least as far as I know of it). But if someone who was researching this (I'm thinking GS, but there may be others), employed two cold readers and two mediums (let's say) to read the same sitters (maybe silent sitter style, maybe not)...with everyone assuming he/she was being read by a "medium"...I think that would be a very interesting way of comparing the readings...focusing on names that were validated (then going on to look at other information to see if there was discrepancy/similarity there or not as well).

The problem, of course, is that even getting names right wouldn't necessarily indicate spirit communication was occurring. But I imagine cold readers would use demographic information so it would be interesting to see how that would stack up vs. mediums' choices.

[Ed]

I might also point out that "analyzing" anything from a TV show is a waste of time.

[Paul C. Anagnostopoulos]

Ed god has spoken verily. Furthermore, why would we trust a written transcript?

Let's ask the mediums to contact Einstein or Feynman and tell us things that the mediums couldn't possibly know. Entire sentences worth of stuff.

~~ Paul

[Ed]

Originally posted by Paul C. Anagnostopoulos
Ed god has spoken verily. Furthermore, why would we trust a written transcript?

Let's ask the mediums to contact Einstein or Feynman and tell us things that the mediums couldn't possibly know. Entire sentences worth of stuff.

~~ Paul

Or German, ad hoc, from Albert.... or anything in any language from anyone as long it is not English. I am sure that there are RULES about this sort of thing.

[T'ai Chi]

Originally posted by CFLarsen
How should we analyze them objectively? Based on what criteria?

Hopefully that is what people will suggest in this thread. Objective meaning we will come to a majority vote of certain interesting things to count or keep track for each transcript.

Letter/name guesses is one area that seems interesting. Of course things like who the medium or cold reader is, if it is a group or a one-on-one session, length in minutes, if a videotape exists, number of animal names used, number of times 'Do you understand that?' or something similar is used, if the transcript review is a skeptic or a believer, but these are just some ideas. I am hoping the thread can come up with more ideas and we could reach some consensus on the good ones to actually keep track of in a study.

[T'ai Chi]

Originally posted by Paul C. Anagnostopoulos
How are you going to calculate the probability that a good hit was simply chance?

~~ Paul

I'm not. Who said I was?

This is simply going to be an observational study where I am only interested in descriptive statistics.

[Ed]

Originally posted by T'ai Chi


I'm not. Who said I was?

This is simply going to be an observational study where I am only interested in descriptive statistics.

Yes, and transcripts of off air shows are meaningless, right?

[CFLarsen]

Originally posted by T'ai Chi
This is simply going to be an observational study where I am only interested in descriptive statistics.

But, how are you going to analyze the transcripts objectively? Please be more specific.

You can't expect people to help you out, if you have no idea what you are trying to achieve.

What about Ed's concern about analyzing transcripts from TV? We know that JE's readings on TV are heavily edited. Why would that give a true picture of what really went on? And what about e.g. Mrs. Piper's transcripts? There is no way we can know for sure if the transcripts are accurate.

How do you plan to deal with all these problems?

[T'ai Chi]

Hey Claus, can you tell Ed "Don't just criticize, come up with better suggestions." too? :)

Ed, I agree that analyzing transcripts is less than ideal, but its all we got. That is why I always push for scientifically studying psi stuff in a lab, and I've also said (in another thread) that videotapes are ideally desired.

Ed, I guess then you must feel that skeptical analysis of JVP or JE transcripts are a waste of time then too.

[CFLarsen]

Originally posted by T'ai Chi
Hey Claus, can you tell Ed "Don't just criticize, come up with better suggestions." too? :)

Ed, I agree that analyzing transcripts is less than ideal, but its all we got. That is why I always push for scientifically studying psi stuff in a lab, and I've also said (in another thread) that videotapes are ideally desired.

Ed, I guess then you must feel that skeptical analysis of JVP or JE transcripts are a waste of time then too.

Whatever. What are you going to do, T'ai Chi? You are going to work with severely flawed data?

[T'ai Chi]

Originally posted by CFLarsen

But, how are you going to analyze the transcripts objectively? Please be more specific.


This is how are we going to anaylze. This won't be a one person job, as I've mentioned.

This thread will hopefully come to a conclusion on interesting and worthwhile things to study in the transcripts. Say, for illustration, that the thread decides on 10 things that seem promising to study in a given transcript. Many transcripts could be gathered and sent to me. Then the transcripts will be randomly assigned and sent out for people to count the 10 things in each transcript. Then they would forward their summaries, the 10*(how-many-transcripts-they-analyze) numbers, back to me. I'd make some graphs showing these descriptive statistics and have the Excel file available for public download.

The objectivity comes in through the counts. No one is hypothesizing anything here, just recording things the thread beforehand decided are worthwhile to record. There might be issues with skeptics giving lower counts to mediums and higher counts to cold readers, or vice versa with believers who analyze the transcripts. However, this would easily be seen by graphic counts for individuals, and constructing side by side box-plots of skeptic counts vs. believer counts. So in short, bias will be out there for all to see if it exists or people decide to get cute.


You can't expect people to help you out, if you have no idea what you are trying to achieve.


As I've already mentioned my idea, it is up to the people to decide if they are interested. If so, great, just PM me if you want to be a transcript gatherer, transcript analyzer, or both. If not, ok, that's fine too, no big deal. I couldn't possibly undertake such a big thing by myself, but I could do something on a smaller scale on my own if there isn't much interest in this.


What about Ed's concern about analyzing transcripts from TV? We know that JE's readings on TV are heavily edited. Why would that give a true picture of what really went on? And what about e.g. Mrs. Piper's transcripts? There is no way we can know for sure if the transcripts are accurate.


Since this is a descriptive study, it is not that big of an issue. We could simply have an indicator variable if the transcript is from a TV show. A side by side box-plot of readings from TV shows and readings not from TV shows would reveal if the TV show reading's counts were different or not.

Likewise with the Piper transcripts, one of the things we could record would be the year in which the reading was done. A reading that is from long ago would have less weight in my opinion, and that's just beacuse we can't be too sure of its accuracy.

[T'ai Chi]

Originally posted by Ersby

That's not going to happen. JE is the foremost medium today, and I've only got about 30 transcripts of him. Asking for hundreds is too tall an order.


Possibly. I did say "ideally" though. :)

Keep in mind that JE is just one medium, but there are many others. There are also some cold readers, whose transcripts need analyzing too.

I think there could be a lot to look over when a lot are gathered up.

[Luke T.]

Well, I suppose you could use some encouragement right about now, T'ai. :D

I noticed your opening post contained the word "interesting," and that should be the accent.

TV transcripts are better than nothing, and if Ersby's John Edward transcripts are the same as what I have seen on tvtalkshow's board, they are pretty good.

Yes, yes, yes, yes, the programs are edited. But if anyone has watched JE's shows like I have, even the best editing can't cover up the fact he sucks. :D

Especially in the last year.

Okay, I'm showing my bias, but who didn't already know it? :)

And Slyvia Browne is a regular on the Montel show. I don't know if he still has her featured every Wednesday or not, but he did not long ago. Simply tape the program and then transcribe it. It's a lot of work, but I think it would be an interesting read nonetheless.

Slyvia's readings suck so bad I can't believe her cover hasn't been blown by the regular exposure. She might as well have a cigarette dangling from her mouth as she does the readings, they are that bad. I always imagine her that way.

And of course there is always Larry King. His shows are live. No editing. He has psychics on all the time. Tape those, too.

We can argue about how to "grade" the "data" some other time. ;)

[CFLarsen]

Originally posted by T'ai Chi
This is how are we going to anaylze. This won't be a one person job, as I've mentioned.

Wait, wait, wait...you want other people to join you in a "a year-long project or longer", but you have no idea yourself how to approach it?

Originally posted by T'ai Chi
This thread will hopefully come to a conclusion on interesting and worthwhile things to study in the transcripts. Say, for illustration, that the thread decides on 10 things that seem promising to study in a given transcript. Many transcripts could be gathered and sent to me. Then the transcripts will be randomly assigned and sent out for people to count the 10 things in each transcript. Then they would forward their summaries, the 10*(how-many-transcripts-they-analyze) numbers, back to me. I'd make some graphs showing these descriptive statistics and have the Excel file available for public download.

The objectivity comes in through the counts. No one is hypothesizing anything here, just recording things the thread beforehand decided are worthwhile to record. There might be issues with skeptics giving lower counts to mediums and higher counts to cold readers, or vice versa with believers who analyze the transcripts. However, this would easily be seen by graphic counts for individuals, and constructing side by side box-plots of skeptic counts vs. believer counts. So in short, bias will be out there for all to see if it exists or people decide to get cute.

But you need to tell us how we count things, how we evaluate things. Please be precise.

Originally posted by T'ai Chi
As I've already mentioned my idea, it is up to the people to decide if they are interested. If so, great, just PM me if you want to be a transcript gatherer, transcript analyzer, or both. If not, ok, that's fine too, no big deal. I couldn't possibly undertake such a big thing by myself, but I could do something on a smaller scale on my own if there isn't much interest in this.

Again, it is not a question of interest. It is a question of you explaining why on earth people should spend so much of their time helping you out, if you have no idea what it is you are trying to do.

Originally posted by T'ai Chi
Since this is a descriptive study, it is not that big of an issue. We could simply have an indicator variable if the transcript is from a TV show. A side by side box-plot of readings from TV shows and readings not from TV shows would reveal if the TV show reading's counts were different or not.

Likewise with the Piper transcripts, one of the things we could record would be the year in which the reading was done. A reading that is from long ago would have less weight in my opinion, and that's just beacuse we can't be too sure of its accuracy.

"Not a big issue"? Are you crazy? We might as well make up transcripts of our own and analyze them!

If data quality is "not a big issue", then I cannot imagine why you would get anyone to help you. Your project is doomed from the start, T'ai Chi, but not from lack of interest. It is doomed because you have no idea how to approach it, and you are willing to work with flawed data.

[Clancie]

Posted by Luke T

And of course there is always Larry King. His shows are live. No editing. He has psychics on all the time. Tape those, too.
LKL actually transcribes them and posts them at the website already. And they're all posted here in the thread where renata analyzed them (maybe we should revisit that thread for starters?)

Maybe they could be a starting point for analyzing other readings, and/or looking at those in a different way. She's already done the "hit/miss" count and it was acceptable to everyone I think.

[CFLarsen]

Originally posted by Luke T.
I noticed your opening post contained the word "interesting," and that should be the accent.

Perhaps. However, a statistical study would be worthless.

Originally posted by Luke T.
And of course there is always Larry King. His shows are live. No editing. He has psychics on all the time. Tape those, too.

He does, and I agree that those transcripts are probably the closest we can get to good data quality. We cannot, however, begin to mix those with transcripts of dubious quality.

Originally posted by Luke T.
We can argue about how to "grade" the "data" some other time. ;)

Nope. We have to be extremely clear about how we are going to deal with the data before we start doing it. Otherwise, we will fare no better than Schwartz: He looks at the data, and goes "Hey, this is significant!"

[T'ai Chi]

Originally posted by CFLarsen

Whatever. What are you going to do, T'ai Chi? You are going to work with severely flawed data?

You can't dismiss all that I said in the last post with a "Whatever."

I already explained what I am going to do. You refuse to ackknowledge it for some odd reason. I'm not losing sleep over your choice.

As far as flawed data, gee, I guess we can't analyze transcripts at all then according to you. I guess skeptics' analyzing of JVP and JE transcripts to demonstrate cold reading is useless. I also guess BillHoyt's "analysis" of the names/letters really is flawed. Thanks for debunking him for me Claus, I appreciate it. Now, back to reality..

Will you follow what you said to Clancie? That is, will you do more than criticize for once and instead offer workable solutions to analyzing transcripts?

If you'd like to, you're welcome to. If, instead you don't want to, I formally invite you to instill doubt and snipe from the sidelines.

I do guarantee that analyzing transcripts will get done with or without your participation.

[CFLarsen]

Originally posted by T'ai Chi
You can't dismiss all that I said in the last post with a "Whatever."

I am merely trying to focus on your suggestion for a big analysis on transcripts. If that is a bad thing, just say so.

Originally posted by T'ai Chi
I already explained what I am going to do. You refuse to ackknowledge it for some odd reason. I'm not losing sleep over your choice.

You have explained it? You will accept working with flawed data? How will you evaluate the transcripts?

Originally posted by T'ai Chi
As far as flawed data, gee, I guess we can't analyze transcripts at all then according to you. I guess skeptics' analyzing of JVP and JE transcripts to demonstrate cold reading is useless. I also guess BillHoyt's "analysis" of the names/letters really is flawed. Thanks for debunking him for me Claus, I appreciate it. Now, back to reality..

Reality, T'ai Chi, is that you willingly accept flawed data.

Originally posted by T'ai Chi
Will you follow what you said to Clancie? That is, will you do more than criticize for once and instead offer workable solutions to analyzing transcripts?

I have most definitely been trying to figure out just what you are trying to do, and how.

Originally posted by T'ai Chi
If you'd like to, you're welcome to. If, instead you don't want to, I formally invite you to instill doubt and snipe from the sidelines.

You may call it "sniping", but I - as well as others - have raised serious doubt about your task here. As for "instilling doubt", I find it sad, but not unexpected, that you want questions about your task to go away. They are not, of course, going to go away.

Originally posted by T'ai Chi
I do guarantee that analyzing transcripts will get done with or without your participation.

Excellent! And I do guarantee that I will be there to evaluate your analysis from a skeptical POV. Heck, I even invite you to publish your findings in SkepticReport!

[T'ai Chi]

Originally posted by CFLarsen

Wait, wait, wait...you want other people to join you in a "a year-long project or longer", but you have no idea yourself how to approach it?


How come you keep saying I have no idea? Please, quit spreading falsehoods. I've mentioned many things that one could study, and mentioned that this thread is about interested people generating more ideas. This tactic to instill doubt from the start is very transparent. I expected better from you, but frankly, you let me down.


But you need to tell us how we count things, how we evaluate things. Please be precise.


You continually try to put this all on me. As I've mentioned, this would be a group effort. Those interested can come to a conclusion on first what to study for a given transcript and then specifically general rules for counting those things.
(so we don't quintuple count letters like Hoyt, for example).


"Not a big issue"? Are you crazy? We might as well make up transcripts of our own and analyze them!


I'm not doing any inference here, just counting things in transcripts and seeing what is out there, as is. Your comparison of making up transcripts of our own and anaylzing them is off the mark. I specifically mentioned I'm interested in mediums and cold readers, ie. JVP, JE, Browne, and other mediums, and Rowland and other cold readers, in readings that have evidence of actually being done (ie. on a tv show? when? where? who was present? what psychical society? etc.)


If data quality is "not a big issue", then I cannot imagine why you would get anyone to help you. Your project is doomed from the start, T'ai Chi, but not from lack of interest. It is doomed because you have no idea how to approach it, and you are willing to work with flawed data.

As I've said, with or without your help this descriptive study will get done. I've also mentioned some ideas, and that this thread is to get more ideas. As it stands, you lead us to believe that you apparently feel that no one can ever analyze any transcripts, although I doubt you have a problem with skeptics analyzing JE and JVP transcripts for cold reading, oddly enough. I wonder if SkepticReport has analyzed any transcripts or readings..
(answer: yes, plenty.
http://www.skepticreport.com/psychics/jeniagara.htm
http://www.skepticreport.com/psychics/jvplkl.htm
http://www.skepticreport.com/psychics/jehitsmisses.htm
http://www.skepticreport.com/psychics/jeblues.htm
http://www.skepticreport.com/astrology/astrologyreadings.htm
http://www.skepticreport.com/psychics/jeomaha.htm
http://www.skepticreport.com/psychics/jelkl.htm
http://www.skepticreport.com/psychics/jespecialhits.htm
http://www.skepticreport.com/psychics/grahambishop.htm
http://www.skepticreport.com/psychics/sblkl.htm)

Interesting.

This position of yours (ie. 'you can't analyze transcripts, but I can!') is worthless for those attempting to understand things because it simply shrugs its shoudlers, puts its hands in the air and says that we can never do anything.

You can keep it. :) I won't be looking at your responses.

[T'ai Chi]

Originally posted by CFLarsen

We have to be extremely clear about how we are going to deal with the data before we start doing it. Otherwise, we will fare no better than Schwartz: He looks at the data, and goes "Hey, this is significant!"

No sh*t re: saying what we will do with the data before we actually do it. That is precisely why this will not be a data mining exercise; because we will, as a group, specify what we will look for and how we will look for it before collecting the data.

[CFLarsen]

Originally posted by T'ai Chi
How come you keep saying I have no idea? Please, quit spreading falsehoods. I've mentioned many things that one could study, and mentioned that this thread is about interested people generating more ideas. This tactic to instill doubt from the start is very transparent. I expected better from you, but frankly, you let me down.

I am so sorry to "let you down". You have still not been very precise: We still have no idea what exactly your aim with this study is.

Originally posted by T'ai Chi
You continually try to put this all on me. As I've mentioned, this would be a group effort. Those interested can come to a conclusion on first what to study for a given transcript and then specifically general rules for counting those things. (so we don't quintuple count letters like Hoyt, for example).

But, surely, you must have had some idea how to approach the subject? How should we evaluate the transcripts?

Originally posted by T'ai Chi
I'm not doing any inference here, just counting things in transcripts and seeing what is out there, as is. Your comparison of making up transcripts of our own and anaylzing them is off the mark. I specifically mentioned I'm interested in mediums and cold readers, ie. JVP, JE, Browne, and other mediums, and Rowland and other cold readers, in readings that have evidence of actually being done (ie. on a tv show? when? where? who was present? what psychical society? etc.)

But you are suggesting an overall analysis of both cold reading transcripts and transcripts from self-proclaimed mediums, yet you want to accept flawed transcripts?

Originally posted by T'ai Chi
As I've said, with or without your help this descriptive study will get done. I've also mentioned some ideas, and that this thread is to get more ideas. As it stands, you lead us to believe that you apparently feel that no one can ever analyze any transcripts, although I doubt you have a problem with skeptics analyzing JE and JVP transcripts for cold reading, oddly enough. I wonder if SkepticReport has analyzed any transcripts or readings..
(answer: yes, plenty.

Thank you for referring to the articles on SkepticReport. However, we are discussing what your analysis can do - or not do. Please stick to that.

Originally posted by T'ai Chi
This position of yours (ie. 'you can't analyze transcripts, but I can!') is worthless for those attempting to understand things because it simply shrugs its shoudlers, puts its hands in the air and says that we can never do anything.

No, I am not saying that. I am merely saying that if you do not have a clear idea of how you will compare the transcripts, your analysis is doomed from the get-go.

Originally posted by T'ai Chi
You can keep it. :) I won't be looking at your responses.

Close your eyes, then. My offer still stands, though: Finish your analysis and I will publish it. Will you submit it, whenever it is finished?

Originally posted by T'ai Chi
No sh*t re: saying what we will do with the data before we actually do it. That is precisely why this will not be a data mining exercise; because we will, as a group, specify what we will look for and how we will look for it before collecting the data.

But please tell us - precisely - how you will do this. You seem to want to heap that responsibility onto others.

How will you work from flawed data, T'ai Chi?

How should we evaluate the transcripts?

What should we look for?

How will you distinguish between a real message from a medium and a lucky guess?

Questions you may ignore, but that doesn't make them go away.

[BillHoyt]

Originally posted by T'ai Chi


Hopefully that is what people will suggest in this thread. Objective meaning we will come to a majority vote of certain interesting things to count or keep track for each transcript.

Letter/name guesses is one area that seems interesting. Of course things like who the medium or cold reader is, if it is a group or a one-on-one session, length in minutes, if a videotape exists, number of animal names used, number of times 'Do you understand that?' or something similar is used, if the transcript review is a skeptic or a believer, but these are just some ideas. I am hoping the thread can come up with more ideas and we could reach some consensus on the good ones to actually keep track of in a study.
The dilettante at work. More amazing, though, might be to ask the dilettante how many such "ideas" he wishes to collect and what alpha level he suggests to set. Oh, that raises a minor issue, doesn't it? Right on the tip of my T'ongue. What was it?

[T'ai Chi]

Originally posted by BillHoyt

The dilettante at work. More amazing, though, might be to ask the dilettante how many such "ideas" he wishes to collect and what alpha level he suggests to set. Oh, that raises a minor issue, doesn't it? Right on the tip of my T'ongue. What was it?

Apparently you know less of statistics than I originally thought (yet that won't stop you from telling me I know nothing funny enough).

I am not going to be doing any inference here, as I've already said several times on this thread. Hence, there is no need to set an alpha.

I won't be reading your respones either. Have a great day. :)

[mgdwcb]

Many psychologists here in the UK such as Dr Richard Wiseman and Dr Chris French say something along the lines of "although there are fraudsters, by enlarge most mediums and psychics are sincere and use cold reading techniques subconsciously." (They may be just patronising the mediums, of course!)

I've looked through the SPR's website and found no material on research on comparing self-confessed cold readers and mediums who insist they are geunune. Its a pity because it would be fascinating. Although there was a study that found that psychics used specific `rhetoric' in their readings as compared to controls (university students)!

[Ersby]

Welcome to the board, mgdwcb. That study using university students: can you remember off hand what it was?

Anyway, I’ve been trying to work out the best way to analyse transcripts for some time. Names are the easiest since we have plenty of name popularity statistics to use. Such that:

“Is your grandfather called John”, would be the probability of a man being called John multiplied by the number of possible targets (two grandfathers). However, this leaves things like “Who’s got the J-o name?” with a probability of being a hit of over one!

I think causes of death could also be looked at in a similar way.

Other things to take into account are the number of people mentioned in the reading. IMO, the more people who “come through” or are talked about, the worse a reading is. Readings should be specific and stay on topic. E.g. If, in the middle of a reading about a school friend who died, the medium asks about “Vicky” and the sitter links it to great aunt Vicky, that should count against the reading.

Repetition of a guess should also be a mark against. Never mind “psychic amnesia”, repeating a guess with subtlety different wording should definitely be a warning sign.

Repetition of subject matter over different readings can also be considered bad. JE likes missing legs, I like to ask about mountains. That kind of thing.

You know, I just remembered. A while ago, I got an email from a student asking about JE, since she was doing a conversational analysis of his work. I’ll email her and ask her how she tackled it.

[Ed]

Originally posted by T'ai Chi
Hey Claus, can you tell Ed "Don't just criticize, come up with better suggestions." too? :)

Practice cold reading so you can be a vampiric bastard too.

Ed, I agree that analyzing transcripts is less than ideal, but its all we got. That is why I always push for scientifically studying psi stuff in a lab, and I've also said (in another thread) that videotapes are ideally desired.

If they were edited, or from an edited source, it is a fool's errand to analyse them and the "results" would mean nothing

Ed, I guess then you must feel that skeptical analysis of JVP or JE transcripts are a waste of time then too.

If they were edited, or from an edited source, it was a fools errand to analyse them and the "results" mean nothing

[mgdwcb]

Originally posted by Ersby
Welcome to the board, mgdwcb.

Thanks.

Originally posted by Ersby
That study using university students: can you remember off hand what it was?

Journal of the Society for Psychical Research, Vol. 64.1, No. 858 " Rheroic in Psychic Detection (Ciaran Ï'Keeffe & Laurence Alison"

The study examined differences between the account-giving styles of `psychic detectives' compared with the control group of students. It was hypothesised that psychics employ devices associated with cold reading style account-giving - psychic rhetoric, they called it. Eight psychics and twelve controls examined 3 objects from 3 crimes and were asked for their opinions about the characteristics of the offender.

Although the psychics were no more accurate than controls, the psychics apparently relied more heavily on `rhetorical devices'.

So it would be interesting to see what differences there are, if any, in `rhetoical devices' between deliberate frauds and those mediums who claim to see and hear dead people.

[BillHoyt]

Originally posted by T'ai Chi
I am not going to be doing any inference here, as I've already said several times on this thread. Hence, there is no need to set an alpha.
Yes, I read it. "Descriptive." Therefore, incapable of reaching any conclusions or of testing any hypotheses. Good choice, dilett'ante.

[CFLarsen]

Originally posted by Ed
If they were edited, or from an edited source, it was a fools errand to analyse them and the "results" mean nothing

Well, yes and no. The interesting thing about the transcripts from an edited source I have seen analyzed and analyzed myself is that even though we know that e.g. JE's readings are even heavily edited, they still show no sign whatsoever of anything else than trickery.

And we know that live readings (LKL) really, really suck. Not just JE, but every medium who tries their hand there.

The sad part of it is that even this will not deter believers. They don't look for weaknesses, but for the odd hit. They don't care that the mediums use tricks or even cheats, as long as the mediums appear - just once in a while - to be talking to dead people.

What we can learn from analyzing these transcripts is just how little it takes to convince the believers. They are desperate to believe.

[Paul C. Anagnostopoulos]

T'ai said:
This is simply going to be an observational study where I am only interested in descriptive statistics.
What are descriptive statistics?

~~ Paul

[BillHoyt]

Originally posted by Paul C. Anagnostopoulos

What are descriptive statistics?

~~ Paul

Mean, mode, standard deviation, etc. They simply describe a population rather than test hypotheses.

[CFLarsen]

Originally posted by BillHoyt
Mean, mode, standard deviation, etc. They simply describe a population rather than test hypotheses.

Descriptive statistics is something a high school student would know about.

[Paul C. Anagnostopoulos]

Those are called descriptive statistics? Hmm. Where have I been all these years? Yes, I guess I've heard that term. I thought he meant statistics about descriptions.

What does it mean to take the mean of a bunch of statements about people's initials? Oh, wait, I know. The mode will be "J" or "M," right? And the standard deviation about B.5?

~~ Paul

[T'ai Chi]

Originally posted by Paul C. Anagnostopoulos

What are descriptive statistics?

~~ Paul

Paul, descriptive statistics are numerical and graphical summaries of samples.

Some examples are the arithmetical mean, geometric mean, harmonic mean, range, 3rd quartile, 1st quartile, interquartile range, median, mode, counts, examining outliers, extreme outliers, standard deviation, skewness, histograms, pie charts, box plots, time series plots, scatterplots, as well as many more, ranging from simple to advanced techniques.

[BillHoyt]

Originally posted by T'ai Chi


Paul, descriptive statistics are numerical and graphical summaries of samples.

Some examples are the arithmetical mean, geometric mean, harmonic mean, range, 3rd quartile, 1st quartile, interquartile range, median, mode, counts, examining outliers, extreme outliers, standard deviation, skewness, histograms, pie charts, box plots, time series plots, scatterplots, as well as many more, ranging from simple to advanced techniques.
Oooh, back at the Waterloo site again, T'ai? But you missed that the charts are not descriptive stats, dilet'tante; they are means of presentation. Why on earth did you choose graphics over such important things as kurtosis? BTW, just because the waterloo text says "skewness" in the first few references doesn't mean the metric is called "skewness." It isn't. It is called "skew." Here, an actual course or textbook, rather than Google, might be your friend.

[Ed]

Originally posted by Paul C. Anagnostopoulos
Those are called descriptive statistics? Hmm. Where have I been all these years? Yes, I guess I've heard that term. I thought he meant statistics about descriptions.

What does it mean to take the mean of a bunch of statements about people's initials? Oh, wait, I know. The mode will be "J" or "M," right? And the standard deviation about B.5?

~~ Paul

B5

Was that a joke or a Freudian slip?

[BillHoyt]

Originally posted by Ed


B5

Was that a joke or a Freudian slip?
C'mon, Ed, it logically follows what went B4.

[T'ai Chi]

Originally posted by BillHoyt

Oooh, back at the Waterloo site again, T'ai? But you missed that the charts are not descriptive stats, dilet'tante; they are means of presentation. Why on earth did you choose graphics over such important things as kurtosis? BTW, just because the waterloo text says "skewness" in the first few references doesn't mean the metric is called "skewness." It isn't. It is called "skew." Here, an actual course or textbook, rather than Google, might be your friend.

(Off ignore for now)

Yeah, I should have said descriptive statistics and descriptive techniques, or descriptive statistics (as numbers) and things that belong in the branch of descriptive Statistics (note the caps), or just Exploratory Data Analysis (EDA) would have done the trick. I didn't though, but I am ever so glad you are here to set us straight. :)

I didn't choose graphics over anything as you erroneously imply. I simply listed many things and also said "as well as many more". You, on the other hand, only listed "Mean, mode, standard deviation, etc." I don't find kurtosis (among other things) on your list either, so please, don't throw stones if you live in a glass outhouse.

By the way, it can, and is, called skewness or skew. Your belief of it only being called skew is very mistaken. http://mathworld.wolfram.com/Skewness.html, and http://www.isixsigma.com/dictionary/Skewness-334.htm, for example. You will also be glad to learn things like means are also called averages, and things like standard deviations are also called spreads. Crazy, I know!!!

Your claims of webpages I've supposedly accessed, without anything other than your own beliefs to support your claim, are empty as when you said that in the past.

I'm glad you mentioned statistics books. Feel free to list all the ones you studied, and then I can ask you questions on what you've mastered. I could start up a thread about statistics for you. ;)

[T'ai Chi]

(repost - Thanks for reminding me Bill and Claus! :) )

What do people think are some valuable things to analyze from the transcripts from self proclaimed mediums and from self proclaimed cold readers?

In this stage I'm building up a list of interesting things to study.

The next stage will be obtaining transcripts, but I need help. I'd like to get hundreds of transcripts ideally, from a variety of mediums and cold readers.

The stage after that will be describing the data (hopefully by assigning a few randomly chosen transcripts to randomly chosen volunteers for them to summarize numerically), using the key things from the first stage. Ideally building a model would be nice, but right now the model part is most definitely overly ambitious. This current project is entirely descriptive in nature.

Please PM me if you are interested in being a volunteer, and please mention whether you would like to help find transcripts, analyze transcripts, or both.
(although please don't start doing these things)

This might be a year-long project or longer, but I think it would be really interesting to undertake if there is enough interest.

[CFLarsen]

T'ai Chi,

So, you are going to ignore the objections and relevant questions put to you here, and simply repeat your previous post (as well as pulling a "Clancie" and taking BillHoyt off ignore right after he says something to you).

Nice going. I see you have learned from Schwartz and Grenard how to conduct an experiment:


Have absolutely no idea what you are looking for.
Start from flawed premises.
Work with flawed data.
Ignore objections.
Data-mine, post hoc.
Inflate the hits, minimize the misses.
Use long, fancy terms that you re-invent the meaning of.
Claim victory, no matter the outcome.


Yup, you're on your way to stardom, T'ai Chi...

[T'ai Chi]

I think I'll respond to you too since you misrepresent me so. I will re-post if you and others keep sniping and provoking.

Originally posted by CFLarsen
T'ai Chi,

So, you are going to ignore the objections and relevant questions put to you here, and simply repeat your previous post (as well as pulling a "Clancie" and taking BillHoyt off ignore right after he says something to you).

Nice going. I see you have learned from Schwartz and Grenard how to conduct an experiment:


Have absolutely no idea what you are looking for.
Start from flawed premises.
Work with flawed data.
Ignore objections.
Data-mine, post hoc.
Inflate the hits, minimize the misses.
Use long, fancy terms that you re-invent the meaning of.
Claim victory, no matter the outcome.


Yup, you're on your way to stardom, T'ai Chi...

None of the items on your list are relevant. I will address them one by one so you don't mispresent me.

I have ideas of what I am looking for which I have shared in this thread, and have mentioned that this is a descriptive study (meaning we collect the data and see what it looks like; we don't do any inference), and have also asked the thread to help think of ideas. This is far from having "absolutely no idea" as you erroneously believe.

What flawed premises? Could you list these premises which you believe I have started from?

I find the 'you can't analyze the data because it is flawed' argument to be empty and hypocritical, given that you, and actual skeptics, have analyzed transcripts a-plenty. The data is out there and I am simply seeing what it looks like.

What objections am I ignoring? So far, I've responded to them all, and found them to be lame. In fact, I've effectively shot these objections to put them out of their misery.

As mentioned before, there is no data mining. Data mining is not having hypotheses, then coming up with hypotheses after examining the data. This is nothing of the sort, as this is a descriptive study and there is no inference taking place. Please ask if you need further clarification on the differences between data mining and looking at descriptive statistics.

I am neither inflating the hits nor minimizing the misses. In fact, I don't even think I've stated that I'd be counting the hits. If the thread decides that hits and misses are sensible things to count, they will both be recorded. Also remember that I'm not going to be the only one counting; far from it. Many believers and many skeptics will be participating if I get more PM's. Also remember, they will be coded which are believers and which are skeptics, so any bias or horseplay, if it exists, will be effortlessly detected for all to see by the simplest of exploratory data analysis methods.

As far those "long, fancy terms that you re-invent the meaning of", could you list any that I have used which are long and fancy and ones I invented? I can only see standard, commonly used and accepted statistical terminology.

And while you're at it, please point out where I claimed victory. I don't even know what you are referring to here.

You can also cease your transparent name-dropping-guilt-by-association tactic of mentioning Clancie, Schwartz, and Grenard when you post to me. But again you miss the point; what I propose isn't an experiment, which tells me you haven't fully grasped the difference between an observational study and an experiment, despite acting pro-science. Egads!

Would you like to offer your useful suggestions for analyzing transcript data? Please, let this thread know. :)

[T'ai Chi]

By the way Claus, any comments on Bill's claim that it is only called skew, or his claim that I get info off of some Waterloo webpage, or him criticizing me for not mentioning kurtosis when he didn't either?

Nah. Thought not. :)

(Yes, I know you'll reply with some pat answer that that is between Bill and I, but that's not true, not when he posts on a bulletin board and his empty arguments are out there for all to see......and critique.)

[CFLarsen]

Originally posted by T'ai Chi
I think I'll respond to you too since you misrepresent me so. I will re-post if you and others keep sniping and provoking.

If you can't stand skeptical scrutiny, you are in the wrong forum.

Originally posted by T'ai Chi
None of the items on your list are relevant. I will address them one by one so you don't mispresent me.

Please read what I write, T'ai Chi: I am describing how Schwartz and Grenard (among others, I should add!) conducts experiments.

My point was to show that you are heading down that road. Fast!

Originally posted by T'ai Chi
I have ideas of what I am looking for which I have shared in this thread, and have mentioned that this is a descriptive study (meaning we collect the data and see what it looks like; we don't do any inference), and have also asked the thread to help think of ideas. This is far from having "absolutely no idea" as you erroneously believe.

"See what it looks like"? But that is exactly my point: You have no idea what you are looking for, you are data-mining, post hoc.

Originally posted by T'ai Chi
What flawed premises? Could you list these premises which you believe I have started from?

Item 1, 3, 4.

Originally posted by T'ai Chi
I find the 'you can't analyze the data because it is flawed' argument to be empty and hypocritical, given that you, and actual skeptics, have analyzed transcripts a-plenty. The data is out there and I am simply seeing what it looks like.

I am pointing to the transcripts of e.g. Piper, and those we cannot verify that it actually happened. I am also pointing to the transcripts which we know are heavily edited.

Are you really going to include transcripts that we cannot check in the slightest? At least we have a chance of verifying what JE said, even though it is heavily edited.

Originally posted by T'ai Chi
What objections am I ignoring? So far, I've responded to them all, and found them to be lame. In fact, I've effectively shot these objections to put them out of their misery.

Finding them "lame" is "ignoring" them, because they have been very relevant to what you are fumbling towards. You have not "effectively shot these objections" by refuting them or pointing out where they were irrelevant, you have simply brushed them off the table. As, as you say, "lame".

Originally posted by T'ai Chi
As mentioned before, there is no data mining. Data mining is nothing having hypotheses, then coming up with hypotheses after examining the data. This is nothing of the sort, as this is a descriptive study and there is no inference taking place. Please ask if you need further clarification on the differences between data mining and looking at descriptive statistics.

How are you going to categorize it? Please don't rely on examples from others, you must have had some idea how you would map the data, when you opened the thread.

Originally posted by T'ai Chi
I am neither inflating the hits nor minimizing the misses. In fact, I don't even think I've stated that I'd be counting the hits. If the thread decides that hits and misses are sensible things to count, they will both be recorded. Also remember that I'm not going to be the only one counting; far from it. Many believers and many skeptics will be participating if I get more PM's. Also remember, they will be coded which are believers and which are skeptics, so any bias or horseplay, if it exists, will be effortlessly detected for all to see by the simplest of exploratory data analysis methods.

But you need to tell us how we are supposed to decide what constitutes a hit or a miss, so we can all see how this bias is dealt with. Or you are going to keep that secret?

Originally posted by T'ai Chi
As far those "long, fancy terms that you re-invent the meaning of", could you list any that I have used which are long and fancy and ones I invented? I can only see standard, commonly used and accepted statistical terminology.

Addressed above: Schwartz, Grenard and others.

Originally posted by T'ai Chi
And while you're at it, please point out where I claimed victory. I don't even know what you are referring to here.

You are doing it as we discuss: You have refused to even consider any objection put to you here. "Lame", remember?

Originally posted by T'ai Chi
You can also cease your transparent name-dropping-guilt-by-association tactic of mentioning Clancie, Schwartz, and Grenard when you post to me. But again you miss the point; what I propose isn't an experiment, which tells me you haven't fully grasped the difference between an observational study and an experiment, despite acting pro-science. Egads!

LOL! Oh, please! I am not employing any "transparent name-dropping-guilt-by-association tactic of mentioning Clancie, Schwartz, and Grenard". I am pointing out that you are doing the same as them.

Originally posted by T'ai Chi
Would you like to offer your useful suggestions for analyzing transcript data? Please, let this thread know. :)

What are your suggestions? This is your thread, you have to come up with something yourself.

Ed is right: The sheer incompetence in paranormal research is staggering...

Originally posted by T'ai Chi
By the way Claus, any comments on Bill's claim that it is only called skew, or his claim that I get info off of some Waterloo webpage, or him criticizing me for not mentioning kurtosis when he didn't either?

Nah. Thought not.

(Yes, I know you'll reply with some pat answer that that is between Bill and I, but that's not true, not when he posts on a bulletin board and his empty arguments are out there for all to see......and critique.)

Wrong again. I have several times explained that I have no education in statistics. You are, of course, free to show yourself that BillHoyt is wrong, instead of just whining that he is...

[T'ai Chi]

Originally posted by CFLarsen

If you can't stand skeptical scrutiny, you are in the wrong forum.


If you can't give proper skeptical scrutiny, you are in the wrong forum.


Please read what I write, T'ai Chi: I am describing how Schwartz and Grenard (among others, I should add!) conducts experiments.

My point was to show that you are heading down that road. Fast!


Since I'm not conducting an experiment your concern is not warranted. As I've already mentioned, if you have ways to improve things, you are invited to give the thread your opinion. You have failed to provide constructive criticism as of yet.


"See what it looks like"? But that is exactly my point: You have no idea what you are looking for, you are data-mining, post hoc.


Your belief is mistaken. Given that data mining is not having hypotheses and then coming up with hypotheses after examining the data, and I am only doing a descriptive study and therefore have no hypotheses because I'm not going to make any inference, there is no data mining. (again)


Item 1, 3, 4.


Regarding items 1, 3, and 4:

1. Again, I have already proposed several things to look for. In addition, again, this thread was created to discuss and agree on things to look for.

3. Your 'you cannot work with flawed data, but I can!' argument was found vacuous long ago.

4. Your objections were demolished in my last few posts.


I am pointing to the transcripts of e.g. Piper, and those we cannot verify that it actually happened. I am also pointing to the transcripts which we know are heavily edited.


I've already mentioned that we can code by year the reading was done and that it would seem sensible to give less weight to these readings. We could also, if the thread thinks sensible, to disregard these few unverifiable readings all together. Also, I find it somewhat interesting, that in the Piper thread, the readings were dismissed as not being very good, etc., but now when one talks of calculating some descriptive statistics possibly using the Piper readings, suddenly things might go all askew because of that. If any readings are very good or not very good or of very high quality or very low quality, these things will stand out once the data is collected, and can be examined. We can examine the 'outliers'.


Finding them "lame" is "ignoring" them, because they have been very relevant to what you are fumbling towards. You have not "effectively shot these objections" by refuting them or pointing out where they were irrelevant, you have simply brushed them off the table. As, as you say, "lame".


No, I have pretty much demolished them, and you are left trying to superglue fragments back together but have only managed to get your hands stickier.


How are you going to categorize it? Please don't rely on examples from others, you must have had some idea how you would map the data, when you opened the thread.


Considering I started the thread to solicit ideas from others, your criticism about "don't rely on examples from others" is irrelevant. Also, as mentioned, I have already mentioned ideas on things to possibly look for.


But you need to tell us how we are supposed to decide what constitutes a hit or a miss, so we can all see how this bias is dealt with. Or you are going to keep that secret?


Again, as I've mentioned, I don't even think I've stated that I'd be counting the hits. If the thread decides that hits and misses are sensible things to count, they will both be recorded, and if that is decided upon, the thread can talk about the specifics of recording hits and misses, and possibly develop some criteria. Personally, as I've mentioned, I think counting hits and misses is too subjective to be carried out.


Addressed above: Schwartz, Grenard and others.


Addressed above: see name-dropping and guilt-by-association tactic. You failed to specifically point out these "long, fancy terms that you re-invent the meaning of" that you believe I use.


You are doing it as we discuss: You have refused to even consider any objection put to you here. "Lame", remember?


As mentioned, I considered them; they just didn't stand up to scrutiny. Again, this is not the same thing as you claiming that I claimed victory.


LOL! Oh, please! I am not employing any "transparent name-dropping-guilt-by-association tactic of mentioning Clancie, Schwartz, and Grenard". I am pointing out that you are doing the same as them.


Your claims of such things have already been addressed and found lacking substance.


What are your suggestions? This is your thread, you have to come up with something yourself.


I have given my suggestions, and I asked for yours. If you have any, you are welcome to present them.


Ed is right: The sheer incompetence in paranormal research is staggering...


His belief (and your belief in him) might be correct. Without evidence, who's to say for certain. Could we see your evidence that there is incompetence so far? I'm glad you call my proposed study research though. Don't let the fact that one can find many examples of sheer incompetence in non-paranormal research sway your beliefs any.


Wrong again. I have several times explained that I have no education in statistics. You are, of course, free to show yourself that BillHoyt is wrong, instead of just whining that he is...

I have conclusively showed that Bill is wrong. What do you conclude from reading his thread on the points I raised (re: Bill's claim that it is only called skew, or his claim that I get info off of some Waterloo webpage, or him criticizing me for not mentioning kurtosis when he didn't either?) and my answer to him? Please, let me know. :)

For someone who claims to have no education in statistics, you sure do seem to have strong opinions about issues directly related to statistics.

[Paul C. Anagnostopoulos]

It's B.5, you guys. Notice the decimal point. I estimate that the standard deviation of guessing initials is about B and one half letters. Come on, stay with the program.

Why doesn't T'ai write up a proposal for the statistical analysis and we can critique it? I'm not sure what descriptive statistics are going to do for us.

~~ Paul

[BillHoyt]

Originally posted by Paul C. Anagnostopoulos
It's B.5, you guys. Notice the decimal point. I estimate that the standard deviation of guessing initials is about B and one half letters. Come on, stay with the program.

Why doesn't T'ai write up a proposal for the statistical analysis and we can critique it? I'm not sure what descriptive statistics are going to do for us.

~~ Paul
In short, Paul, they do little of anything. It is T'ai's science-as-dilletante-stamp-collecting. But, in that spirit, I thought we might look at an interesting collection of information:
<table border="1" width="90%">
<tr>
<td width="50%">T'ai's Or'der</td>
<td width="50%">Waterloo¹</td>
</tr>
<tr>
<td width="50%"></td>
<td width="50%"></td>
</tr>
<tr>
<td width="50%">arithmetical mean</td>
<td width="50%">arithmetic mean</td>
</tr>
<tr>
<td width="50%">geometric mean</td>
<td width="50%">harmonic mean</td>
</tr>
<tr>
<td width="50%">harmonic mean</td>
<td width="50%">geometric mean</td>
</tr>
<tr>
<td width="50%">range</td>
<td width="50%">median</td>
</tr>
<tr>
<td width="50%">3rd quartile</td>
<td width="50%">mode</td>
</tr>
<tr>
<td width="50%">1st quartile</td>
<td width="50%">range</td>
</tr>
<tr>
<td width="50%">interquartile range</td>
<td width="50%">1st quartile</td>
</tr>
<tr>
<td width="50%">median</td>
<td width="50%">3rd quartile</td>
</tr>
<tr>
<td width="50%">mode</td>
<td width="50%">interquartile range</td>
</tr>
<tr>
<td width="50%">counts</td>
<td width="50%">standard deviation</td>
</tr>
<tr>
<td width="50%">examining outliers</td>
<td width="50%">skewness</td>
</tr>
<tr>
<td width="50%">extreme outliers </td>
<td width="50%">examining outliers</td>
</tr>
<tr>
<td width="50%">standard deviation</td>
<td width="50%">box plots</td>
</tr>
<tr>
<td width="50%">skewness</td>
<td width="50%">histograms</td>
</tr>
<tr>
<td width="50%">histograms</td>
<td width="50%">pie charts</td>
</tr>
<tr>
<td width="50%">pie charts</td>
<td width="50%">scatterplots</td>
</tr>
<tr>
<td width="50%">box plots</td>
<td width="50%"></td>
</tr>
<tr>
<td width="50%">time series plots</td>
<td width="50%"></td>
</tr>
<tr>
<td width="50%">scatterplots</td>
<td width="50%"></td>
</tr>
</table>

____
¹
Waterloo (http://www.science.uwaterloo.ca/course_notes/biology/biol361/lecture04.ppt)

[T'ai Chi]

Originally posted by Paul C. Anagnostopoulos
It's B.5, you guys. Notice the decimal point. I estimate that the standard deviation of guessing initials is about B and one half letters. Come on, stay with the program.

Why doesn't T'ai write up a proposal for the statistical analysis and we can critique it? I'm not sure what descriptive statistics are going to do for us.

~~ Paul

Descriptive statistics tell us what the data is like as it is 'found in the wild'. So far, to the best of my knowledge, people haven't looked at a large number of transcripts and/or kept detailed numerical records of their findings. This makes it difficult for people who make claims of what medium or cold reader transcripts supposedly contain, or the relationships between medium and cold reader transcripts. Only by getting actual data can we get a clue as to those things.

[CFLarsen]

T'ai Chi,

You seem more interested in "demolishing" critique instead of actually getting on with the job. You are not interested in doing "descriptive statistics", so far all you have done is trying to raise more and more questions (e.g. asking for evidence of incompetence in paranormal research, when you are perfectly aware of many of the examples). You also seem much more interested in getting "even" with me, BillHoyt and whoever has pissed you off.

Whatever it is you are trying to do has nothing to do with the subject of the thread. You are only a troll in a sandbox.

Happy research. Nothing will come of it, of course, except you whining about us not doing your job for you.

Originally posted by Paul C. Anagnostopoulos
Why doesn't T'ai write up a proposal for the statistical analysis and we can critique it? I'm not sure what descriptive statistics are going to do for us.

A.k.a. "calling the bluff". Yes, T'ai Chi, why don't you, instead of all this hand-waving? Focus on whatever it is you are trying to do.

Added:

Originally posted by T'ai Chi
Descriptive statistics tell us what the data is like as it is 'found in the wild'. So far, to the best of my knowledge, people haven't looked at a large number of transcripts and/or kept detailed numerical records of their findings. This makes it difficult for people who make claims of what medium or cold reader transcripts supposedly contain, or the relationships between medium and cold reader transcripts. Only by getting actual data can we get a clue as to those things.

Bluff called. T'ai Chi folds. No statistical analysis.

[T'ai Chi]

Originally posted by BillHoyt

In short, Paul, they do little of anything.


Descriptive statistics do little of anything? Really? They seem to do something of importance, especially when desiring to see what the actual data looks like. :) For you to attempt to persude me (and others- that is the despicable part) that descriptive statistics "do little of anything", when in reality they are the tools for describing and exploring samples, and used all the time in many disciplines, especially science, is a total joke on your part, and I seriously question your understanding, specifically of statistics, and science in general.


It is T'ai's science-as-dilletante-stamp-collecting.


Your buzzwords are interesting, but if you could explain with rational, statistical terms, that would be more productive. So far, this 'tse tse woo fly' has spanked you from here to there and back again. Looks like you need to call some more arms.


But, in that spirit, I thought we might look at an interesting collection of information:


Ok... so you believe that that is evidence for me copying my list from Waterloo?? Perhaps you could give a model and calculate a probability for us?

Your table is fine on my part, but you not so suprisingly left out some things on theirs. They mentioned the midrange, variance, coefficient of variation, standard error of the mean, confidence intervals, kurtosis, and bar charts (as distinct from histograms). Besides the lists already looking fairly different, they are even more different when one factors in the things you consciously or unconsciously omitted. It doesn't even look like you can read a paper and make list correctly.

While we're on the topic of statistics,
Here's an interesting one for you:

Independent posts by Bill, each of which contains intelligent thoughts with probability p = .05, are performed until there are 10 consecutive posts with intelligent thoughts. What is the mean number of necessary posts by Bill?

[CFLarsen]

T'ai Chi,

Drop the personal wars. Get on with your descriptive statistics instead.

That is what you want to do, isn't it?

[T'ai Chi]

Originally posted by CFLarsen
T'ai Chi,

You seem more interested in "demolishing" critique instead of actually getting on with the job.


Should I be impressed at your attempts at mind-reading? I'm simply responding to you, Bill, and Paul, that's all.


You are not interested in doing "descriptive statistics", so far all you have done is trying to raise more and more questions (e.g. asking for evidence of incompetence in paranormal research, when you are perfectly aware of many of the examples).


I am interested in doing descriptive statistics, hence the reason for this thread, and hence the reason why your critique is unwarranted.


You also seem much more interested in getting "even" with me, BillHoyt and whoever has pissed you off.


I've simply addressed and corrected your, and his, misrepresentation of the facts.


You are only a troll in a sandbox.


Possibly. If you could address some statistical things, that would be great. Again, you are invited to offer your constructive criticism on how to analyze medium transcripts. I am aware you and others on SkepticReport have analyzed this "flawed data" before, so maybe you could awe me with your expertise.


Happy research. Nothing will come of it, of course, except you whining about us not doing your job for you.


Thanks! As I said, such a study will get done with or without you; no big loss.


Bluff called. T'ai Chi folds. No statistical analysis.

Of course there is no statistical analysis, because there is no data collected yet. Brilliant Claus, really.

[CFLarsen]

T'ai Chi,

Very well: You've addressed our points. Time to get on with your descriptive statistics now.

[BillHoyt]

Originally posted by T'ai Chi
Ok... so you believe that that is evidence for me copying my list from Waterloo?? Perhaps you could give a model and calculate a probability for us?
Thank you, T'ai, for falling into the trap. I said I was doing it in the spirit of your science-as-stamp-collecting here. That is, not imposing models, not trying to draw conclusions, just gathering up the data. But I see you seem to be bothered that, in this context, no evidence can be provided by such stamp collecting. I see you now saying a "model" needs to be here. Hmmm.

Quod erat dumbandstrandum.

[Ed]

Originally posted by T'ai Chi


Descriptive statistics tell us what the data is like as it is 'found in the wild'. So far, to the best of my knowledge, people haven't looked at a large number of transcripts and/or kept detailed numerical records of their findings. This makes it difficult for people who make claims of what medium or cold reader transcripts supposedly contain, or the relationships between medium and cold reader transcripts. Only by getting actual data can we get a clue as to those things.

Except that when there is any editing by anyone with any bias, the data is diefinitively not "in the wild" it has been domesticated to a greater or lesser degree and such an anlysis is useless.

[CFLarsen]

Originally posted by Ed
Except that when there is any editing by anyone with any bias, the data is diefinitively not "in the wild" it has been domesticated to a greater or lesser degree and such an anlysis is useless.

This point is continuously ignored by T'ai Chi. Why is that?

[T'ai Chi]

Originally posted by CFLarsen
T'ai Chi,

Very well: You've addressed our points. Time to get on with your descriptive statistics now.

I'm glad you conceeded the first point.

Also, thank you for your concern, but I've already been getting on with thinking about the specifics of the study, and communicating with people who show interest.

Again, you are welcome to offer your constructive criticism on how to analyze medium transcripts.

[T'ai Chi]

Originally posted by BillHoyt

Thank you, T'ai, for falling into the trap.


LOL, what, this is a game to you? Trolling are we? :D


I said I was doing it in the spirit of your science-as-stamp-collecting here. That is, not imposing models, not trying to draw conclusions, just gathering up the data.


That's not science as stamp collecting, that is describing and exploring samples, something which is used in every science. If you have a way other than statistics to do that, I'm all ears. I, for one, don't know what transcript data looks like in general. As far as I am aware, no one has attempted to study it numerically in any great detail.


But I see you seem to be bothered that, in this context, no evidence can be provided by such stamp collecting. I see you now saying a "model" needs to be here. Hmmm.


No, you just need to back up your claims of me supposedly copying from a website. I haven't made any claims about cold reading transcripts or medium transcripts. Surely you can make an effort to understand the difference.

I also noticed you ignored this:

Here's an interesting one for you:

Independent posts by Bill, each of which contains intelligent thoughts with probability p = .05, are performed until there are 10 consecutive posts with intelligent thoughts. What is the mean number of necessary posts by Bill?

Care to take a stab? :)

[CFLarsen]

Originally posted by T'ai Chi
I'm glad you conceeded the first point.

Still getting "even", eh?

Originally posted by T'ai Chi
Also, thank you for your concern, but I've already been getting on with thinking about the specifics of the study, and communicating with people who show interest.

Of course, this is secret. Naturally.

Originally posted by T'ai Chi
Again, you are welcome to offer your constructive criticism on how to analyze medium transcripts.

Let's see the actual analyses first. If they ever show...

[T'ai Chi]

Originally posted by Ed

Except that when there is any editing by anyone with any bias, the data is diefinitively not "in the wild" it has been domesticated to a greater or lesser degree and such an anlysis is useless.

Possibly, but I find it hypocritical that skeptics can analyze transcripts and claim to get valuable information showing cold reading has taking place, among other things, but when someone else wants to analyze the data the data is incredibly flawed, biased, edited, etc.

As mentioned, if the transcript is from a TV reading, that could be noted in the spreadsheet. One could construct a box plot of TV reading transcript vs. non-TV reading transcripts and see if there are any obvious differences, for example.

This is about seeing what the data looks like, not doing drawing conclusions about that data to some population of cold reader and medium transcripts.

[T'ai Chi]

Originally posted by CFLarsen

This point is continuously ignored by T'ai Chi. Why is that?

Semi-interesting belief of yours, but I've addressed that point more than once.

[CFLarsen]

T'ai Chi,

Whatever. How far have you gotten with your descriptive statistical analysis?

[T'ai Chi]

Originally posted by CFLarsen

Still getting "even", eh?


Not at all. I was simply pointing out the fact that you conceeded that I've addressed your and your comrades' points.


Of course, this is secret. Naturally.


Only because you haven't PM'd me, as instructed, if you are interested.


Let's see the actual analyses first. If they ever show...

You seem to have read things backwards. I welcomed your constructive criticism on how to analyze transcripts, and you asked for the analyses first. :rolleyes: Of course, you are welcome to criticize after the fact. If that occurs (which is fine), I will certainly be there to point out that you did not offer your constructive criticism before the fact though.

After things are hammered out, the full dataset will be made public as an Excel file for download. As I already mentioned, due to the large number of transcripts out there (and that fact that I and other people involved have jobs and lives outside of here), I think this will be a project that will take a few years. I, for one, really only have time on the weekends, and am not planning spending every weekend doing this. :)

As I've said, I don't think anybody has numerically analyzed large quantities of transcripts before, so it should be fairly useful to people who are interested in the areas of cold reading, mediumship, skepticism, etc.

[CFLarsen]

Originally posted by T'ai Chi
blah, blah, blah...

You have no idea how to approach this.

You have no idea how to evaluate the data.

You have no idea when this is ready for publication.

Like I said: Schwartz, Grenard...and now you.

Let a trembling world know when you are done.

[T'ai Chi]

Originally posted by CFLarsen
T'ai Chi,

Whatever. How far have you gotten with your descriptive statistical analysis?

You cannot dismiss what I said with a "Whatever". That is not acceptable rational discourse.

Thank you for asking how I am doing so far, although it seems you are interested in progress but not offering your a priori constructive criticism. Oh well. As for me, I've offered, in this thread, some possible things I thought of to analyze for any given transcript, completely addressed your and your comrades' points regarding unfounded criticisms, have heard some interesting things from people who are interested in it, and set up some columns reflecting these possible variables in an Excel spreadsheet. Considering this study is only 2 or 3 days young, it is off to a promising start, and I thank you for your help and interest. :)

[T'ai Chi]

Originally posted by CFLarsen

You have no idea how to approach this.

You have no idea how to evaluate the data.

You have no idea when this is ready for publication.

Like I said: Schwartz, Grenard...and now you.

Let a trembling world know when you are done.

Thank you for your interest, but I did not say blah blah blah.

You have failed to rationally address several points which I have put to you.

I have addresed your above unfounded criticisms completely and conclusively for all to see. Thanks again. :)

[CFLarsen]

T'ai Chi,

And I will be checking on you regularly to see how your work is going.

Question is: Will you show how far you've come?

[TLN]

Originally posted by T'ai Chi


Thank you for your interest, but I did not say blah blah blah.

You have failed to rationally address several points which I have put to you.

I have addresed your above unfounded criticisms completely and conclusively for all to see. Thanks again. :)

T'ai Chi, show us your work or shut the f#ck up. This playground nonsense of yours is getting really tiresome.

Put up or shut up. It's not up to us to do your work for you. Do it yourself and stop taunting others like it's their responsibility to do it for you.

[T'ai Chi]

Originally posted by CFLarsen
T'ai Chi,

And I will be checking on you regularly to see how your work is going.


I'm glad to hear it. I would also be glad to hear, as I've mentioned, your constructive criticism on how to analyze medium transcripts since you have done it before and posted articles by others who have done it and seem to be an expert in skeptical thinking and cold reading..


Question is: Will you show how far you've come?

I will PM those who have PM'd me showing their interest. Once the study actually gets in full swing, communication will be done by PM's or email, not in threads here. If you'd like to be in the loop and offer constructive criticism, just PM me and let me know.

[T'ai Chi]

Originally posted by TLN

T'ai Chi, show us your work or shut the f#ck up. This playground nonsense of yours is getting really tiresome.

Put up or shut up. It's not up to us to do your work for you. Do it yourself and stop taunting others like it's their responsibility to do it for you.

Pardon me TLN, but I think you are off the mark here.

I mentioned what has been done over the last two days regarding planning and the Excel file. If you'd like to be in the loop, as mentioned, feel free to PM me. If you complain because you aren't, despite my instructions regarding PM-ing, I can only assume you consciously or unconsciously overlooked my writing, despite me mentioning it several times.

I've already mentioned, numerous times, that this descriptive study is a group effort, so your "It's not up to us to do your work for you" idea is again, off the mark, as I am not the only one doing the work. I've simply asked for your, Claus's, or anyone else's constructive comments on how to analyze transcripts.

That is, the purpose of this thread is seeking input and will take suggestions on things to keep track of for a given transcript. You are welcome to participate as usual. :)

[CFLarsen]

Originally posted by T'ai Chi
I'm glad to hear it. I would also be glad to hear, as I've mentioned, your constructive criticism on how to analyze medium transcripts since you have done it before and posted articles by others who have done it and seem to be an expert in skeptical thinking and cold reading..

It is interesting that you seem more preoccupied with fighting your personal wars instead of actually focusing on the things you say you want to do.

Originally posted by T'ai Chi
I will PM those who have PM'd me showing their interest. Once the study actually gets in full swing, communication will be done by PM's or email, not in threads here. If you'd like to be in the loop and offer constructive criticism, just PM me and let me know.

I am not clear on why this has to be kept secret? Why not inform everyone on how it is going? If you can inform one, why not all?

[TLN]

Originally posted by T'ai Chi
I mentioned what has been done over the last two days regarding planning and the Excel file. If you'd like to be in the loop, as mentioned, feel free to PM me. If you complain because you aren't, despite my instructions regarding PM-ing, I can only assume you consciously or unconsciously overlooked my writing, despite me mentioning it several times.

Why is it this paranormal clap-trap is always done in secret? Because you know you don't have a clue how to proceed, so you're hoping to taunt Claus into telling you how to proceed. What's next? Calling him "chicken" and flapping your wings? Go back to the playground.

Originally posted by T'ai Chi
I've already mentioned, numerous times, that this descriptive study is a group effort, so your "It's not up to us to do your work for you" idea is again, off the mark, as I am not the only one doing the work. I've simply asked for your, Claus's, or anyone else's constructive comments on how to analyze transcripts.

Request denied. Get it? Now go away. I'm sure you'll manage fine on your own, won't you?

I look forward to the results of your solo effort. {holds breath}

[BillHoyt]

T'ai,

I'll save you the effort of typing your future posts.

<marquee>Bl'ah bl'ah b'lah bla'h wh'atever bl'ah bla'h bla'h bla'ther bl'ah b'lah.</marquee>

You can go home now. Either that, or stop this puerile posing.

[Ed]

Originally posted by T'ai Chi


Possibly, but I find it hypocritical that skeptics can analyze transcripts and claim to get valuable information showing cold reading has taking place, among other things, but when someone else wants to analyze the data the data is incredibly flawed, biased, edited, etc.

As mentioned, if the transcript is from a TV reading, that could be noted in the spreadsheet. One could construct a box plot of TV reading transcript vs. non-TV reading transcripts and see if there are any obvious differences, for example.

This is about seeing what the data looks like, not doing drawing conclusions about that data to some population of cold reader and medium transcripts.

I never made any such claim. I would not waste my time either analizing such transcripts nor reading about such analyses. T'ai, it does not matter how considered or diliberate you sound, if the data is suspect anything that flows from it is questionable. Proceeding with lousey data, with descriptive stats or anything else is only valid in describing a flawed data set. It is on a par with Ian's silliness. It signifies nothing.

The fact that you refuse to address this except to point out another bad example does nothing to justify looking at junk. It also suggests your veriage about science and analysis is rather specious.

[T'ai Chi]

Originally posted by CFLarsen

I am not clear on why this has to be kept secret? Why not inform everyone on how it is going? If you can inform one, why not all?

Those who have PM'd me, as instructed, who are directly involved with the study, will be notified. You are welcome to do your own study where you can do things that you think sensible.

By the way, I found this today: http://tvtalkshows.com/forum/showthread.php?threadid=27301&highlight=Cantata. You tell me that analyzing cold reading and medium transcripts is not wise because the data is "flawed", yet you did that very thing.

In the future, please apply your criteria to all people, not just those you have a personal beef with. :)

Thanks.

[T'ai Chi]

Originally posted by TLN

Why is it this paranormal clap-trap is always done in secret? Because you know you don't have a clue how to proceed, so you're hoping to taunt Claus into telling you how to proceed. What's next? Calling him "chicken" and flapping your wings? Go back to the playground.


It won't be done in secret, but it will be done by those who have PM'd me with their interest, as instructed to. You seem to ignore that part. As far as having a clue, as mentioned several times, I already listed my ideas about possible things to look for for a given transcript, and created this thread to solicit suggestions on things to look for.


Request denied. Get it? Now go away. I'm sure you'll manage fine on your own, won't you?


As I've already said, with or without you or Claus or Ed or Bill, this study will get done. I'm not managing on my own, but rather, as I've already mentioned, will have many people assisting with the counts.


I look forward to the results of your solo effort. {holds breath}

Great, I am happy that you are interested. :) Again, you are welcome to offer your constructive comments on what to look for in a given transcript.

[T'ai Chi]

Originally posted by BillHoyt
T'ai,

I'll save you the effort of typing your future posts.

<marquee>Bl'ah bl'ah b'lah bla'h wh'atever bl'ah bla'h bla'h bla'ther bl'ah b'lah.</marquee>

You can go home now. Either that, or stop this puerile posing.

Unfortunately (for your trolling efforts), I'm not going anywhere and neither is this project. Seeing what the data actually is and studying, numerically, using descriptive statistics, a large number of cold reader and medium transcripts is very useful. It will get done with or without your assistance as I've already mentioned.

Again, as I've already mentioned, you are welcome to offer your constructive comments on what to look for in a given transcript.

I will also note that you have also failed to address my response of:


No, you just need to back up your claims of me supposedly copying from a website. I haven't made any claims about cold reading transcripts or medium transcripts. Surely you can make an effort to understand the difference.


So I will consider that as you conceeding that point of mine and that I am correct that you cannot actually back up your claim.

You have also failed to address your mistaken claim that it is only called skew, when clearly skewness is used often. I'm really not surprised at your behavior regarding your claims.

By the way, the answer to :

"Independent posts by Bill, each of which contains intelligent thoughts with probability p = .05, are performed until there are 10 consecutive posts with intelligent thoughts. What is the mean number of necessary posts by Bill?"

is 10,778,947,368,420 posts. Maybe my estimate of p was a little too low, but again, maybe not.

[T'ai Chi]

Originally posted by Ed

I never made any such claim.


You might not have, sure.


T'ai, it does not matter how considered or diliberate you sound, if the data is suspect anything that flows from it is questionable.


Then would you agree that skeptics analyzing transcripts for cold reading is questionable?

Since I'm not doing inference, I'm merely describing the data that is out there, it isn't a big deal. As mentioned, we can create a column in the spreadsheet that records if the transcript is from a TV reading, or a live reading, or from a book, etc. If we make a graph of some counts, for example, and we notice that the ones coded 'TV reading' are higher, for example, then this reveals to us that in the data we studied there is something special with the TV readings that could be explained by sensible things like editing, for example. We won't know until we get the actual data. Again, keep in mind I've pretty much decided to not look at subjective things like hits or misses.


Proceeding with lousey data, with descriptive stats or anything else is only valid in describing a flawed data set. It is on a par with Ian's silliness. It signifies nothing.


You have yet to really convince that it is unworthy of study and that edited = flawed. According to you, one can never analyze any transcript, anything on TV, anything in book form, anything in any type of media (except completely live recordings/movies) which is surely a worthless way of looking at things because nothing ever gets done. I disagree with it 100%, for reasons I mentioned above.


The fact that you refuse to address this except to point out another bad example does nothing to justify looking at junk. It also suggests your veriage about science and analysis is rather specious.

Considering I've addressed your point and found it unconvincing and hypocritical, your above points are rather vacuous.

[BillHoyt]

Originally posted by T'ai Chi


Unfortunately (for your trolling efforts),

Aww, h'*ll fo'lks, it is just more bl'ather from h'im. The only one who is a bit third moment around here is our p'oser. Welcome to ig'nore, you third derivative.

[CFLarsen]

Originally posted by T'ai Chi
Those who have PM'd me, as instructed, who are directly involved with the study, will be notified. You are welcome to do your own study where you can do things that you think sensible.

You have still not addressed the question: Why does this have to be secret?

Originally posted by T'ai Chi
By the way, I found this today: http://tvtalkshows.com/forum/showthread.php?threadid=27301&highlight=Cantata. You tell me that analyzing cold reading and medium transcripts is not wise because the data is "flawed", yet you did that very thing.

In the future, please apply your criteria to all people, not just those you have a personal beef with. :)

Thanks.

I don't give a flying f*ck what you have found. Please address this question: Why does your analysis have to be secret?

Originally posted by T'ai Chi
It won't be done in secret, but it will be done by those who have PM'd me with their interest, as instructed to.

Why? Why not inform everyone??

Originally posted by T'ai Chi
As I've already said, with or without you or Claus or Ed or Bill, this study will get done. I'm not managing on my own, but rather, as I've already mentioned, will have many people assisting with the counts.

People, who are not named, and the data not identified. It is all done in secrecy. Why?

Originally posted by T'ai Chi
Seeing what the data actually is and studying, numerically, using descriptive statistics, a large number of cold reader and medium transcripts is very useful. It will get done with or without your assistance as I've already mentioned.

So, why are this not done in public? Why done in secrecy?

Originally posted by T'ai Chi
"Independent posts by Bill, each of which contains intelligent thoughts with probability p = .05, are performed until there are 10 consecutive posts with intelligent thoughts. What is the mean number of necessary posts by Bill?"

is 10,778,947,368,420 posts. Maybe my estimate of p was a little too low, but again, maybe not.

Please keep focused here.

Why do you need to do this analysis in secrecy?

[TLN]

Originally posted by T'ai Chi
It won't be done in secret, but it will be done by those who have PM'd me with their interest, as instructed to. You seem to ignore that part.

Ignore it? That's the whole friggin' point!

PM = in secrecy.

So, I'll ask you again, Claus style so you can be sure to ignore it like the worthless troll you are:

Why do you need to do this analysis in secrecy?

[Ed]

Originally posted by T'ai Chi


You might not have, sure.

As I said....

Then would you agree that skeptics analyzing transcripts for cold reading is questionable?

Questionable? You purport to know something about analysis. Such an analysis is self referential. It tells you about the data you analysed and nothing more. Not just transcripts, transcripts of edited broadcasts.

Since I'm not doing inference, I'm merely describing the data that is out there, it isn't a big deal.

Of course not. It is just not meaningful and any "results" are not generalizable to anything. This is sorta why someone elses home movies are deadly dull. Let us assume all Pam Anderson and Paris Hilton jokes are told so that we can move on.

As mentioned, we can create a column in the spreadsheet that records if the transcript is from a TV reading, or a live reading, or from a book, etc. If we make a graph of some counts, for example, and we notice that the ones coded 'TV reading' are higher, for example, then this reveals to us that in the data we studied there is something special with the TV readings that could be explained by sensible things like editing, for example.

It may tell you something, you have no idea what the cause is. Editing confounds everything. Also, different editors might have different subjective rules. You just do not know and anything you say is speculation.

We won't know until we get the actual data. Again, keep in mind I've pretty much decided to not look at subjective things like hits or misses.

Sorry, subjective? Objective, surely.

You have yet to really convince that it is unworthy of study and that edited = flawed.

Edited=Edited. You have no idea what the "sampling distribution" looks like. The data, due to the editing process is suspect. In real science suspect=Flawed. It is another form of lack of control.

According to you, one can never analyze any transcript, anything on TV, anything in book form, anything in any type of media (except completely live recordings/movies) which is surely a worthless way of looking at things because nothing ever gets done.

Where, precisely, did I say that? Without you being more specific, I really can't comment on books and other things but if you had an accurate transcript of an event from start to finish I would be reasonably comfortable ... providing that you had a couple of cases to demonstrate the integrity of the findings. Remember that "any type of media" would preclude any data collection whatsoever.

I disagree with it 100%, for reasons I mentioned above.

I don't see reasons beyond willful insistance. By all means analyse away. Realize that no one will take results from such data seriously and this does not reflect close minded scepticism simply rationality. Note that I am poining out the problems with your approach prior to seeing any results, pro or con. Science, you see, does not wait to see results before method is passed on.

Considering I've addressed your point and found it unconvincing and hypocritical, your above points are rather vacuous.

You have not really addressed anything except to state that you do not comprehend the problem. Fair. Might you describe why I am hypocritical? I would really like an answer to that, or an apology. As for my points being vacuous, since I am fairly sure that you do not really understand what it is that you are doing you will pardon me for taking that slight with a grain of salt. I will let the readers here judge whether my points have merit or not.

[BillHoyt]

Originally posted by Ed


My Ed, Ed! For Ed's sake, how'd you miss T'ai's slip out of descriptive stat into inference testing? Here it is again:

"As mentioned, we can create a column in the spreadsheet that records if the transcript is from a TV reading, or a live reading, or from a book, etc. If we make a graph of some counts, for example, and we notice that the ones coded 'TV reading' are higher, for example, then this reveals to us that in the data we studied there is something special with the TV readings that could be explained by sensible things like editing, for example."

It reveals nothing to us because we started with no hypothesis. It reveals nothing to us because we have determined no way to distinguish normal random fluctuations from significant ones. Ed, T'ai is a complete p'oser. He can't even get these basics down, no matter how much googling he does.

This reveals nothing to us. Particularly when we start out data mining, then try to impose alpha levels and then claim significance. This was Claus' point early on that the p'oser was too wh'atever to understand.

P'oser is a third derivative.

[Ed]

Originally posted by BillHoyt


My Ed, Ed! For Ed's sake, how'd you miss T'ai's slip out of descriptive stat into inference testing? Here it is again:

"As mentioned, we can create a column in the spreadsheet that records if the transcript is from a TV reading, or a live reading, or from a book, etc. If we make a graph of some counts, for example, and we notice that the ones coded 'TV reading' are higher, for example, then this reveals to us that in the data we studied there is something special with the TV readings that could be explained by sensible things like editing, for example."

It reveals nothing to us because we started with no hypothesis. It reveals nothing to us because we have determined no way to distinguish normal random fluctuations from significant ones. Ed, T'ai is a complete p'oser. He can't even get these basics down, no matter how much googling he does.

This reveals nothing to us. Particularly when we start out data mining, then try to impose alpha levels and then claim significance. This was Claus' point early on that the p'oser was too wh'atever to understand.

P'oser is a third derivative.

I guess I just sorta felt like writing. But you are correct, I believe. The matter of fact recitation of a litany, been laden with buzzwords, is a giveaway. I have found that smart people, not brilliant necessarily, just smart, if they know their stuff tend to adapt real time. Walk away from positions, float positions to get information then modify them and so on. Without any wish to be insulting, it seems to me that both T'ai and Ian have a sctick that is oiled and practiced and which succeeds more often then not in their personal lives. When they get pushed you get T'ai insistance or Ian denseness. Both defences I suppose.

[T'ai Chi]

Originally posted by BillHoyt

Aww, h'*ll fo'lks, it is just more bl'ather from h'im. The only one who is a bit third moment around here is our p'oser. Welcome to ig'nore, you third derivative.

Again, Bill fails to rationally address things.

Glad to see you cracked open that calculus book, and yes I get it. I did think my problem was a little more imaginative though.

[T'ai Chi]

Originally posted by CFLarsen

You have still not addressed the question: Why does this have to be secret?


If you are interested in helping, you can PM me as I've mentioned. Also, feel free to offer your expert advice on how to analyze transcripts, since you've done that in the past.

Are you made aware of the progress of every study that is currently going on in any discipline? No. Gee, why all of that secrecy? :rolleyes:

As I've already stated, the data set will be made available when the study is completed. As stated, the study will not be conducted through posting to threads, just the idea gathering and logistics stages.

If you disagree, feel free to start your own study. :)

[T'ai Chi]

Originally posted by TLN


Ignore it? That's the whole friggin' point!

PM = in secrecy.

So, I'll ask you again, Claus style so you can be sure to ignore it like the worthless troll you are:

Why do you need to do this analysis in secrecy?

Your attempted and emotional tag-team antics are most amusing. :)

Carry on.

OR, feel free to offer your constructive comments on how to analyze transcripts.

[CFLarsen]

T'ai Chi,

Very well, then:

Why do you need to do this analysis in secrecy?

[T'ai Chi]

Ed, you wrote:

"...any "results" are not generalizable to anything."

You hit the nail on the head, but are slightly wrong. They are not generalizable to the larger population of cold reader and medium transcripts (because of non-randomness of their selection and it is an observational study) but that doesn't mean the results aren't useful, as they surely would surely tell us interesting information about the sample cold reader and medium transcripts.

For example, how many 'ethnic' names have been used in cold reader or medium readings as judged by the count of 'ethnic' names in cold reader or medium transcripts? No one can answer questions like these without looking at the actual data.

[T'ai Chi]

Originally posted by BillHoyt

then this reveals to us that in the data we studied there is something special with the TV readings that could be explained by sensible things like editing, for example."


That's not a change from descriptive to inferential statistics dude. Note the words "in the data we studied". Try again. :)


Ed, T'ai is a complete p'oser. He can't even get these basics down, no matter how much googling he does.


So you continually believe and claim but have utterly failed to demonstrate.


P'oser is a third derivative.

You have failed to admit your errors, such as skewness being a commonly used and accepted word, despite your ranting otherwise, for example.

[TLN]

Originally posted by T'ai Chi
Your attempted and emotional tag-team antics are most amusing. :)

And your evasion of a very simple question is most amusing.

Science isn't done in secret, Troll Chi. It's done in the light of day for everyone to comment on. Obviously, you're a graduate of the Gary Schwartz school of obstruction.

Please stop bothing us with your obvious trollish baiting and playground antics.

Grow up.

[CFLarsen]

TLN,

One reason why T'ai Chi wants to keep it secret:

He has already started creating false impressions.

[T'ai Chi]

Originally posted by TLN

And your evasion of a very simple question is most amusing.


If by "evasion" you mean that I have answered it, then I agree.


Science isn't done in secret, Troll Chi. It's done in the light of day for everyone to comment on. Obviously, you're a graduate of the Gary Schwartz school of obstruction.


Are you made aware of the updates from every scientific study out there? No? Why not TLN? Why don't all those scientists inform you about it by sending out an email to you everytime they change their experimental protocols or add another row to their spreadsheet? :)


Please stop bothing us with your obvious trollish baiting and playground antics.

Grow up.

If you have constructive comments, you are more than welcome to submit them. A simple way for me to stop "bothing" you is for you to put me on ignore.

[T'ai Chi]

Originally posted by CFLarsen
TLN,

One reason why T'ai Chi wants to keep it secret:

He has already started creating false impressions.

Everything I said was factual and without exxageration.

You'll have to do better Claus.

All Ersby could offer was his own transcripts. If he is a professional cold reader, then this is suitable. If he is not, it is not. If no one can point to cold reader transcripts, I find it odd that people can make statements about them with any confidence. For example, I doubt anyone can even answer the most basic question of how many J-names are guessed in a large number of cold reader transcripts, among others.

[CFLarsen]

Answered here. (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870357924#post1870357924)

[T'ai Chi]

Here are the column headers of my Excel spreadsheet so far (with brief explanation in parentheses). These are some things that I and others have thought would be sensible to examine for any given transcript.

Any constructive comments on other things to examine or things to omit are most welcome.

Transcript Finders (list of people who volunteered to find transcripts for the study, for reference)

Analyzers (list of people who volunteered to examine transcripts for the study, for reference)

TranscriptID (a unique number, 1 through # of transcripts in the study, is assigned to each transcript, for reference)

WhereFound (the exact webpage, book, etc. where the transcript was found)

WhoFound (the person who found the transcript)

WhoAnalyzed (the person who examined the transcript)

WhoDoubleChecked (the person who doublechecked the original examination of the transcript by WhoAnalyzed, if necessary if there is any serious debate over the counts)

NameGuessCount (a count of the total names (first and last) guessed by the medium or cold reader in a transcript)

LETTERS GUESSED---> (the following are counts of the letters guessed by the medium or cold reader for a given transcript, ie: "I'm getting an R" counts as 1.)
a
b
c
d
e
f
g
h
I
j
k
l
m
n
o
p
q
r
s
t
u
v
w
x
y
z

EthnicNameCount (the count of ethnic names guessed for a transcript)

EthnicName (the ethnic name(s) guessed in a transcript, for future reference)

Where (the venue where the readings in the transcript were done, ie. on LKL for example)

Group setting/1:1 in person/Calls (type of readings done in the transcript)

Cold Reader/Medium (is the person a self proclaimed cold reader or a self proclaimed medium?)

Name of Cold Reader/Medium (name of cold reader or medium, ie JVP, Rowland, etc.)

DateOfReading (the date the reading was given in month/day/year format)

OnTV (were the readings in the transcript done on tv, yes or no)

Live? (were the readings in the transcript done on live TV, yes or no)

#DoYouUnderstandThats (the number of 'do you understand that?' in a transcript)

#ReadingsPerTranscript (the number of readings in a transcript)

SavedTranscriptOnHardDrive? (did I save the transcript on my harddrive? These should all be Yes. It is really there to remind myself to make sure that I do that.)

[T'ai Chi]

Originally posted by CFLarsen
Answered here. (http://www.randi.org/vbulletin/showthread.php?s=&postid=1870357924#post1870357924)

If you believe that, I have a bridge to sell you.

[CFLarsen]

Originally posted by T'ai Chi
Here are the column headers of my Excel spreadsheet so far (with brief explanation in parentheses). These are some things that I and others have thought would be sensible to examine for any given transcript.

Any constructive comments on other things to examine or things to omit are most welcome.

Interesting to see that you have abandoned the secrecy. Why did you want to keep it secret?

Originally posted by T'ai Chi
Transcript Finders (list of people who volunteered to find transcripts for the study, for reference)

Analyzers (list of people who volunteered to examine transcripts for the study, for reference)

TranscriptID (a unique number, 1 through # of transcripts in the study, is assigned to each transcript, for reference)

WhereFound (the exact webpage, book, etc. where the transcript was found)

WhoFound (the person who found the transcript)

WhoAnalyzed (the person who examined the transcript)

WhoDoubleChecked (the person who doublechecked the original examination of the transcript by WhoAnalyzed, if necessary if there is any serious debate over the counts)

NameGuessCount (a count of the total names (first and last) guessed by the medium or cold reader in a transcript)

Where can we check the actual transcripts? Will they be stored for viewing?

Originally posted by T'ai Chi
LETTERS GUESSED---> (the following are counts of the letters guessed by the medium or cold reader for a given transcript, ie: "I'm getting an R" counts as 1.)

What about phonetics? "John", "George"..."I'm getting a "DJ-sound".

Originally posted by T'ai Chi
a
b
..
y
z

What about names that begin with æ, ø, å? You need to include all possible letters.

Originally posted by T'ai Chi
EthnicNameCount (the count of ethnic names guessed for a transcript)[/B]

Define "ethnic name". What is that?

Originally posted by T'ai Chi
Group setting/1:1 in person/Calls (type of readings done in the transcript)[/B]

What about those readings by JE, where he first does them as a group, then hauls them on stage?

Originally posted by T'ai Chi
Cold Reader/Medium (is the person a self proclaimed cold reader or a self proclaimed medium?)[/B]

So, you have given up on your demand that the cold reader has to be "professional"? Why?

Originally posted by T'ai Chi
OnTV (were the readings in the transcript done on tv, yes or no)

What about other media? Radio, Internet?

Originally posted by T'ai Chi
Live? (were the readings in the transcript done on live TV, yes or no)

What about other media? Radio, Internet?

Originally posted by T'ai Chi
#DoYouUnderstandThats (the number of 'do you understand that?' in a transcript)

What about all the other techniques, e.g. those mentioned in Rowland's book?

Originally posted by T'ai Chi
#ReadingsPerTranscript (the number of readings in a transcript)

You don't separate the different readings??

Originally posted by T'ai Chi
SavedTranscriptOnHardDrive? (did I save the transcript on my harddrive? These should all be Yes. It is really there to remind myself to make sure that I do that.)

They have to be available for everyone. Will they be?

What about ethnicity/cultural background? You have to include every possible ethnicity and culture, along with the most common names/letters.

[Ed]

You did not address one point that I made.

[BillHoyt]

Originally posted by Ed
You did not address one point that I made.
I hope the omniscient Ed is not surprised by the behavior of one of his children. Ya buy 'em books and ya buy 'em books and all they do is chew on the covers.

[Ed]

Remarkable. Also a good indication of the state of paranormal research.

[BillHoyt]

Originally posted by Ed
Remarkable. Also a good indication of the state of paranormal research.

I think here we have a special case, although I'm not sure. T'ai's lack of math and stat skills, and Grenard's, should probably not be used as an indicator. But you are right that the fact that researchers can't keep it straight, coupled with the fact that some of them use people like T'ai and Grenard as front men may be indicative.

One thing I find sad is that they don't know they are nothing but trial balloons. The Grenards and Bolens and others of the world are encouraged to keep skeptics busy with their bizarre assertions and to keep the masses guessing. If they get shot down, so what? There was no direct connection to those researchers. They are summarily popped, and left to float away. That they don't get their actual position in the scheme of things makes them rather "trial buffoons."

We'll need to watch what becomes of this study. The need for secrecy is interesting, as is the slip into inferential statistics. That T'ai has already spoken of collecting reams of data in secret and then also spoken about drawing conclusions from that data says clearly that he is data mining. He will collect up dozens of sets of data, analyze them in secret and report significance at the .05 level. No dice, as anybody skilled in statistics would know.

"As mentioned, we can create a column in the spreadsheet that records if the transcript is from a TV reading, or a live reading, or from a book, etc. If we make a graph of some counts, for example, and we notice that the ones coded 'TV reading' are higher, for example, then this reveals to us that in the data we studied there is something special with the TV readings that could be explained by sensible things like editing, for example."

This is an amazingly revealing quote! T'ai is now trying to walk away from it And this is an amazingly revealing attempt at a defense:

"That's not a change from descriptive to inferential statistics dude. Note the words "in the data we studied". "

The words T'ai cites underscore my point. It is exactly a switch to inferential statistics. This is the same error T'ai committed while discussing triple-blind studies. You cannot choose your model or your alpha or your hypothesis after the data have been collected. You cannot cherry-pick your data set and report it as significant unless you report all the other data you have gone through. Period. End of story. The significance is gone when you play those games.

[Paul C. Anagnostopoulos]

Surely you're not planning to analyze scripts from TV shows?

Where are the following columns:

AmazingHits
PofAmazingHitsbyChance
CountofItemsGivenbySitterThenUsedLater
CountofThinkAboutThisatHome
CountofTimesPsychicHuntedforSitter
CountofDoesThisMakeSense
CountofI'mSensingtheHeadorChestArea
CountofYesThat'sWhatI'mGetting

~~ Paul

[Ersby]

I'd have to say all of T'ai Chi's criteria are superficial and meaningless. Nothing suggested has anything to do with the actual mechanics of a reading. A simple count of letter frequency won't let you know if cold reading/genuine mediumship is occurring. I thought we'd already gone through that. And whether something is on TV or not is beside the point. Only "Do you understand that?" may be worth counting.

I gave some input on what to examine on page one of this thread.

[BillHoyt]

Originally posted by Ersby
I'd have to say all of T'ai Chi's criteria are superficial and meaningless. Nothing suggested has anything to do with the actual mechanics of a reading. A simple count of letter frequency won't let you know if cold reading/genuine mediumship is occurring. I thought we'd already gone through that. And whether something is on TV or not is beside the point. Only "Do you understand that?" may be worth counting.

I gave some input on what to examine on page one of this thread.

There was a wave of "literary criticism" around in the 60s and 70s. Perhaps it still exists? Very scientific; used computers. The people doing it scanned great works into the computer and had the computer spit out word counts. 2,700 "the"; 4,356 "a"; 52 "rump". That was it. Just word counts. They even published books containing these counts. Pages and pages of word counts. Very useful. Very very useful. Hello? You can wake up now.

[T'ai Chi]

Originally posted by BillHoyt

I hope the omniscient Ed is not surprised by the behavior of one of his children. Ya buy 'em books and ya buy 'em books and all they do is chew on the covers.

Bill, could you admit that you were wrong by saying it is only called skew?

Thanks expert.

[T'ai Chi]

Originally posted by Ersby
I'd have to say all of T'ai Chi's criteria are superficial and meaningless. Nothing suggested has anything to do with the actual mechanics of a reading.


If you have any further constructive comments on alternatives, feel free to chime in. This is the thread for doing that. :)


A simple count of letter frequency won't let you know if cold reading/genuine mediumship is occurring. I thought we'd already gone through that.


I thought I had already gone through that this study isn't seeking to determine that. This is merely a descriptive study.

[T'ai Chi]

Originally posted by Paul C. Anagnostopoulos
Surely you're not planning to analyze scripts from TV shows?


Why not? As mentioned, I have a column indicating if it was done on TV.


Where are the following columns:

AmazingHits
PofAmazingHitsbyChance
CountofItemsGivenbySitterThenUsedLater
CountofThinkAboutThisatHome
CountofTimesPsychicHuntedforSitter
CountofDoesThisMakeSense
CountofI'mSensingtheHeadorChestArea
CountofYesThat'sWhatI'mGetting

~~ Paul

As mentioned, this study will probably not be counting hits, because it tends to get more subjective than I'd like. The third and especially the fourth item in your list interest me, as well as your seventh item.

[Clancie]

re: the badgering T'ai over "Why do your research in secrecy?"...:rolleyes: ....I will quote T'ai's comment from another thread:
Posted by T'ai Chi

As far as so-called secret research, I suppose that Claus gets updates from all researchers all over the world whenever they change even the most minor thing in their study.
Yes, T'ai, for them to be badgering you about this is about the stupidest thing I've ever seen on this board.

And, btw, Bill. Weren't you just the one criticizing people here for wasting bandwidth with insults and content-less threads where the badgering and baiting (as opposed to real discussion) just dragged on and on?

I can't help wondering when you are going to start contributing to those discussions you say JREF so desperately needs--you know, the ones where you set such a great example for us all of what intellectual dialogue and educational content in a thread really look like?
:i:

Please. Seriously. Don't forget to let everyone know when you decide to start practicing what you preach.

[BillHoyt]

Originally posted by BillHoyt


There was a wave of "literary criticism" around in the 60s and 70s. Perhaps it still exists? Very scientific; used computers. The people doing it scanned great works into the computer and had the computer spit out word counts. 2,700 "the"; 4,356 "a"; 52 "rump". That was it. Just word counts. They even published books containing these counts. Pages and pages of word counts. Very useful. Very very useful. Hello? You can wake up now.

60s : 1
70s : 1
a : 1
and : 2
around : 1
book: : 1
can : 1
computer : 3
count : 3
criticism : 1
even : 1
exist : 1
hello : 1
in : 1
it : 2
just : 1
now : 1
of : 2
literary : 1
pages : 2
perhaps : 1
published: 1
scientific : 1
that : 1
the : 1
there : 1
they : 2
up : 1
very : 4
wake : 1
was : 1
wave : 1


Very useful.

[T'ai Chi]

Originally posted by BillHoyt

T'ai's lack of math and stat skills,


You have yet to convince anyone. You were so convinced that the word is only skew. Give me a break dude. :)


them use people like T'ai and Grenard as front men may be indicative.


haha, lol. Is this some conspiracy that I don't know about??


One thing I find sad is that they don't know they are nothing but trial balloons. The Grenards and Bolens and others of the world are encouraged to keep skeptics busy with their bizarre assertions and to keep the masses guessing. If they get shot down, so what? There was no direct connection to those researchers. They are summarily popped, and left to float away. That they don't get their actual position in the scheme of things makes them rather "trial buffoons."


We sure do have amazing properties! We are apparently a rare product of evolution; magical tse tse balloon flies.

"encouraged to keep skeptics busy".. lol, Bill, you are truly on the fringe of being a conspiracy theory believer.


We'll need to watch what becomes of this study.


I'm glad. :) As I mentioned several times, if you have any constructive comments on what to look for in transcripts, let this thread know about it. I don't want you to whine later that this and that wasn't done, when you have been given the opportunity to constructively chime in plenty of times..


The need for secrecy is interesting,


You, along with Claus, must have a flooded mail and email box from all the updates that are pouring in from researchers all across the world!


as is the slip into inferential statistics.


As already mentioned, you made a claim of me supposedly copying, one that you haven't backed up. I'm not making any claims about medium or cold reader transcripts, I'm just recording counts and various other things from their transcripts.


That T'ai has already spoken of collecting reams of data in secret and then also spoken about drawing conclusions from that data says clearly that he is data mining.


As I've mentioned, data mining is when you have no hypotheses, collect data, look at the data for hypotheses to test, then test that same data. I am simply collecting data and seeing what this data looks like. I'm not doing any inference, no matter how strongly you believe I am.


He will collect up dozens of sets of data, analyze them in secret and report significance at the .05 level. No dice, as anybody skilled in statistics would know.


Was your attempt at predicting the future supposed to impress anybody?


This is an amazingly revealing quote! T'ai is now trying to walk away from it And this is an amazingly revealing attempt at a defense:


Yes it is.

Note the words "in the data we studied". Descriptive statistics tell us about the data we studied.

[b]
You cannot choose your model or your alpha or your hypothesis after the data have been collected.
[/;b]

Uh, where did I choose an alpha or a model here? Maybe you could actually back up your claims?

Thanks. :)

[CFLarsen]

T'ai Chi,

Why does this have to be secret?

[BillHoyt]

Originally posted by CFLarsen
T'ai Chi,

Why does this have to be secret?

<marquee> Councellor! Stop badgering the wooness!</marquee>

[Clancie]

<marquee> Bill! "Counselor" is spelled with an "s" after the "n" ! Learn to spell better if you're going to insult others!</marquee>

[T'ai Chi]

Originally posted by CFLarsen

Interesting to see that you have abandoned the secrecy. Why did you want to keep it secret?


Major updates, after the project is in full-swing after logisitics are hammered out, are given to those who are participating in the study. You can start your own study if you disagree and would like to structure it differently.


Where can we check the actual transcripts? Will they be stored for viewing?


As mentioned, the exact reference to the transcript will be recorded, as well as a copy of the transcript being saved to my hard drive. People can go to the exact reference and save the transcipts to their own hard drive if they'd like.


What about phonetics? "John", "George"..."I'm getting a "DJ-sound".


I plan to count "John, George" as 2 names guessed. I'm not sure how to record things like "DJ-sound". I am welcome to hearing your suggestions?


What about names that begin with æ, ø, å? You need to include all possible letters.


Those could get recorded too in the odd chance that they arise, sure.


Define "ethnic name". What is that?


One that doesn't fit with the typical pattern of names in the population where the reading is taking place. For example, say a medium is reading a caucasian American person in the USA and says 'Who is Qin Xi?". I believe Ed was asking about ethnic names in another thread and I thought would be interesting to record.


What about those readings by JE, where he first does them as a group, then hauls them on stage?


Those might be difficult to record. I could record, instead of Group, 1:1, or Calls, them as 1, 2, and 3. If it is first Group and then 1:1 I could record a (1+2)/2 = 1.5, for example. Do you have any suggestions?


So, you have given up on your demand that the cold reader has to be "professional"? Why?


No, I'm still only interested in study transcripts of professionals in this study.


What about other media? Radio, Internet?


Those would not be TV, so a N would go in the OnTV column. One could record, specifically, the type of media, but I just had it dichotimized on TV.


What about all the other techniques, e.g. those mentioned in Rowland's book?


Feel free to actually mention them. :)



You don't separate the different readings??


One could. Do you feel that is important? I'm studying transcripts as the
unit, not the reading. If you have an argument for studying readings as the unit, feel free to present your thoughts.


They have to be available for everyone. Will they be?


As mentioned the exact references will be given. I've also will save copies of every transcript on my hard drive. If a person is interested, they can follow the exact reference and save them to their own hard drive.


What about ethnicity/cultural background? You have to include every possible ethnicity and culture,

You'll have to present a more persuasive argument than "You have to.." for me to consider it. Personally, including every ethnicity and culture is silly as there are many, many thousands of them.

[CFLarsen]

Originally posted by Clancie
<marquee> Bill! "Counselor" is spelled with an "s" after the "n" ! Learn to spell better if you're going to insult others!</marquee>

It is not bad form to point out spelling errors, but it is bad form to mock because of spelling errors.

Anything to get back at those you don't like, eh?

[T'ai Chi]

Originally posted by BillHoyt


60s : 1
70s : 1
a : 1
and : 2
around : 1
book: : 1
can : 1
computer : 3
count : 3
criticism : 1
even : 1
exist : 1
hello : 1
in : 1
it : 2
just : 1
now : 1
of : 2
literary : 1
pages : 2
perhaps : 1
published: 1
scientific : 1
that : 1
the : 1
there : 1
they : 2
up : 1
very : 4
wake : 1
was : 1
wave : 1


Very useful.


In the way you are strawmanning it, no, it's not. :) Although, you believe that descriptive statistics don't do much, so I'm not surprised.

Your view has some skewness.

[T'ai Chi]

Originally posted by CFLarsen
T'ai Chi,

Why does this have to be secret?

Already addressed.

You must have a flooded mail and email box from all the updates that are pouring in from researchers all across the world! :)

The exact reference for the transcripts will be recorded, and a copy of each transcript will be saved to my hard drive. Interested people can go to the exact reference and save the transcripts to their own hard drive. People involved with the study, after the logistics and specifics are hammered out in the thread will get updates. Everyone at the end of the study will have access to download the full Excel spreadsheet. Everyone now has the opportunity to offer constructive comments on what to look for in transcripts.

[T'ai Chi]

Originally posted by BillHoyt

<marquee> Councellor! Stop badgering the wooness!</marquee>

Bill, you're not swaying this woo-woo tse tse fly balloon in the slightest. You might need to call in some more arms.

I'm glad I'm on your mind though.

Bill, is it only called skew or not? Why do you have so hard a time being honest and admitting you were talking out of your butt? :)

[CFLarsen]

Originally posted by T'ai Chi
Major updates, after the project is in full-swing after logisitics are hammered out, are given to those who are participating in the study. You can start your own study if you disagree and would like to structure it differently.

But why only to those participating? Why not everyone? You don't want to answer this, do you?

Originally posted by T'ai Chi
As mentioned, the exact reference to the transcript will be recorded, as well as a copy of the transcript being saved to my hard drive. People can go to the exact reference and save the transcipts to their own hard drive if they'd like.

So, nobody will ever see what data you used. Grrrrrrrreat, T'ai Chi!

Originally posted by T'ai Chi
I plan to count "John, George" as 2 names guessed. I'm not sure how to record things like "DJ-sound". I am welcome to hearing your suggestions?

You have to map out the possible phonetic combinations. Yes, you. You have to do work yourself.

Originally posted by T'ai Chi
Those could get recorded too in the odd chance that they arise, sure.

How do you know anything about chance, when you haven't even looked at the data?

Originally posted by T'ai Chi
One that doesn't fit with the typical pattern of names in the population where the reading is taking place. For example, say a medium is reading a caucasian American person in the USA and says 'Who is Qin Xi?". I believe Ed was asking about ethnic names in another thread and I thought would be interesting to record.

How are you going to determine what the typical pattern of names in the population where the reading is taking place is? How are you going to determine what ethnicity a sitter has?

Would you consider a name like "Juanita", "Jose" and "Carlos" "ethnic" names? Why/why not?

Originally posted by T'ai Chi
Those might be difficult to record. I could record, instead of Group, 1:1, or Calls, them as 1, 2, and 3. If it is first Group and then 1:1 I could record a (1+2)/2 = 1.5, for example. Do you have any suggestions?

No. What are you going to do with those readings?

Originally posted by T'ai Chi
No, I'm still only interested in study transcripts of professionals in this study.

Then you have moved the goal posts. And lied about it, too.

Originally posted by T'ai Chi
Those would not be TV, so a N would go in the OnTV column. One could record, specifically, the type of media, but I just had it dichotimized on TV.

Why only TV?

Originally posted by T'ai Chi
Feel free to actually mention them. :)

Read the book, T'ai Chi. You have to do something yourself.

Originally posted by T'ai Chi
One could. Do you feel that is important? I'm studying transcripts as the
unit, not the reading. If you have an argument for studying readings as the unit, feel free to present your thoughts.

You can't even see the problem? JE reads person 1, a hispanic. Ethnic names are expected. Then, in the same transcript, he reads person 2, an Irish. Ethnic names are unexpected.

You really don't see a problem with that??

Originally posted by T'ai Chi
As mentioned the exact references will be given. I've also will save copies of every transcript on my hard drive. If a person is interested, they can follow the exact reference and save them to their own hard drive.

So, you would only include what is online? What if a transcript suddenline disappears, because of a broken link? Have you thought of that?

Originally posted by T'ai Chi
You'll have to present a more persuasive argument than "You have to.." for me to consider it. Personally, including every ethnicity and culture is silly as there are many, many thousands of them.

(groan) If you don't, what are you going to use the ethnicity for??

You are utterly incompetent, T'ai Chi.

Originally posted by T'ai Chi
In the way you are strawmanning it, no, it's not. Although, you believe that descriptive statistics don't do much, so I'm not surprised.

Please point out what is different from your approach.

[T'ai Chi]

Originally posted by CFLarsen

But why only to those participating? Why not everyone? You don't want to answer this, do you?


They are the ones involved in the study.


So, nobody will ever see what data you used. Grrrrrrrreat, T'ai Chi!


Huh? They will see the exact transcripts.


You have to map out the possible phonetic combinations. Yes, you. You have to do work yourself.


I'm asking you for your suggestions not you to do my work. Why are you so hostile? Why do you refuse to offer constructive comments?


How do you know anything about chance, when you haven't even looked at the data?


From transcripts I've seen in the past. If you have evidence that a lot of those letters and names are used, please feel free to present it.


How are you going to determine what the typical pattern of names in the population where the reading is taking place is?


Statistics are available to everyone at the Census Bureau.


How are you going to determine what ethnicity a sitter has?


I'm not.. it is the ethnicity the name guessed by the medium or cold reader.


Would you consider a name like "Juanita", "Jose" and "Carlos" "ethnic" names? Why/why not?


No, those are pretty common names in the US, for example. In China, they would be more uncommon. :)


No. What are you going to do with those readings?


You don't have any constructive comments on how to record those? Interesting. Well, I offered a way to record them, so I'll probably record them the way I suggested unless anybody has more worthwhile ideas.


Why only TV?


I dichotmized it on TV as those are the readings most available. Why do you feel it should have been dichotomized somewhere else?


Read the book, T'ai Chi. You have to do something yourself.


I'm not asking you to do work, I'm asking for your constructive suggestions. It amazes me that you confuse these very different things.


You really don't see a problem with that??


It is up to you to reveal your reasoning why you see it as a problem.


So, you would only include what is online? What if a transcript suddenline disappears, because of a broken link? Have you thought of that?


Not only include. References can be in any form, webpages, books, etc. If a transcripts suddenly disappears, because of a broken link, efforts will be made to find the correct link. Regardless, copies of transcripts will be saved.


You are utterly incompetent, T'ai Chi.


If this belief of yours makes you feel better, by all means, keep believing it. You have failed to present any evidence for this. In any case, I am specifically asking people for their constructive ideas on how to examine transcripts, so if I have left out something, which I am not beyond doing, I am specifically hoping someone will see it and offer useful suggestions.



In short, I'm not incompetent because you refuse to collect data or offer constructive ideas on how to examine transcripts.

But, this coming from someone who says it is worthless to analyze transcripts because of the flawed data, yet freely analyzes his own and puts up articles of people who have analyzed transcripts is quite amusing. Will you address this point ever Claus?

Please point out what is different from your approach.

I simply desire to look at more meaningful things. I'm not counting words like "the", "in", and "it" for example. I'm also not scanning in things and having a computer count words. I''m also not going to be the only one doing it; believers and skeptics, anyone, can participate. I'm also not doing literary analysis from the 60's and 70's. The fact that Bill equates descriptive statistics with this is quite revealing of his convoluted thinking and his lame bulletin board debate tactics.

[CFLarsen]

Originally posted by T'ai Chi
They are the ones involved in the study.

Yes, that is obvious. But why only these people? If they can know, why not all?

Originally posted by T'ai Chi
Huh? They will see the exact transcripts.

Perhaps the people involved. However, everything is not made public, is it?

Originally posted by T'ai Chi
I'm asking you for your suggestions not you to do my work. Why are you so hostile? Why do you refuse to offer constructive comments?

I have note "refused" to offer constructive comments. I have made a lot, OK? But whenever something that looks like real work comes up, you insist that others do it for you.

Originally posted by T'ai Chi
From transcripts I've seen in the past. If you have evidence that a lot of those letters and names are used, please feel free to present it.

What transcripts are those? How did you analyze them statistically?

Originally posted by T'ai Chi
Statistics are available to everyone at the Census Bureau.

How are you going to determine the typical pattern of names in the population at a Crossing Over-reading? People come from all over the country, don't they?

Originally posted by T'ai Chi
I'm not.. it is the ethnicity the name guessed by the medium or cold reader.

Say again?

Originally posted by T'ai Chi
No, those are pretty common names in the US, for example. In China, they would be more uncommon. :)

How are you going to determine what names are "ethnic"?

Originally posted by T'ai Chi
You don't have any constructive comments on how to record those? Interesting. Well, I offered a way to record them, so I'll probably record them the way I suggested unless anybody has more worthwhile ideas.

Your way is flawed. Are you still going to go on with that? Sure you are...

Originally posted by T'ai Chi
I dichotmized it on TV as those are the readings most available. Why do you feel it should have been dichotomized somewhere else?

That way, you are lumping unedited readings from LKL together with edited readings from CO. You don't see a problem with that?

Originally posted by T'ai Chi
I'm not asking you to do work, I'm asking for your constructive suggestions. It amazes me that you confuse these very different things.

I am not confusing anything. I gave you the book reference, now you do the work yourself.

Originally posted by T'ai Chi
It is up to you to reveal your reasoning why you see it as a problem.

If you mix the ethnicities within the same transcript, you mix the results. Understand?

Originally posted by T'ai Chi
Not only include. References can be in any form, webpages, books, etc. If a transcripts suddenly disappears, because of a broken link, efforts will be made to find the correct link. Regardless, copies of transcripts will be saved.

Which makes the links irrelevant. You can therefore make all transcripts publically available. Will you?

Originally posted by T'ai Chi
If this belief of yours makes you feel better, by all means, keep believing it. You have failed to present any evidence for this. In any case, I am specifically asking people for their constructive ideas on how to examine transcripts, so if I have left out something, which I am not beyond doing, I am specifically hoping someone will see it and offer useful suggestions.

Again, you merely brush off any criticism by referring to it as "belief".

Originally posted by T'ai Chi
But, this coming from someone who says it is worthless to analyze transcripts because of the flawed data, yet freely analyzes his own and puts up articles of people who have analyzed transcripts is quite amusing. Will you address this point ever Claus?

Please keep focused. We are discussing your flawed analysis here.

Originally posted by T'ai Chi
I simply desire to look at more meaningful things. I'm not counting words like "the", "in", and "it" for example. I'm also not scanning in things and having a computer count words. I''m also not going to be the only one doing it; believers and skeptics, anyone, can participate. I'm also not doing literary analysis from the 60's and 70's. The fact that Bill equates descriptive statistics with this is quite revealing of his convoluted thinking and his lame bulletin board debate tactics.

So, what are you doing, and for what purpose?

[T'ai Chi]

Originally posted by CFLarsen

Yes, that is obvious. But why only these people? If they can know, why not all?


As mentioned those are the people directly involved with the study. It doesn't make sense, logisitcally and logically, to email all people regarding a study. All people who want to know can know, if they PM me, as stated, saying they would like to be involved in the study, as mentioned. So you have no excuse for claiming that things are secret, as you've been informed many times on how to be updated once the project is in full swing.

I'm wondering if you could you give examples of studies that have made all of their results available to the all the public at every step? Please, amaze us with examples.. You seem to be holding this to a higher standard for some bizarre reason.

You can also start up your own study and do things differently if you'd like.


But whenever something that looks like real work comes up, you insist that others do it for you.


I have not insisted that others do my work. I have asked for ideas on what people would examine in a given transcript though. Personally, I've already worked about 2 hours on thinking, setting up a spreadsheet, and incorporating peoples' suggestions into it. Replying to scoffers that call me an incompetent tse tse fly and balloon only occupies a very uninteresting 20 minutes of that time.


What transcripts are those? How did you analyze them statistically?


Just transcripts in general. I didn't keep a record of those which had names that started with those characters though. At that time I wasn't keeping records for one, and for another, I don't recall any of them ever had any such names. I could be wrong though, I'm always open to that evidence, which reminds me:

Perhaps you could find examples of transcripts that do and convince this thread why worry about such things is important as you believe, but you might think that is me asking for you to do my work.


How are you going to determine the typical pattern of names in the population at a Crossing Over-reading? People come from all over the country, don't they?


No, the typical pattern of names where the reading is being done. Since they are mostly readings in the US, the Census Bureau has statistics on names, demographics, etc. No problem there. :)


How are you going to determine what names are "ethnic"?


If the name is not a common name in the location in which the reading is being done. For example, if the readers says he's getting a John Smith but the reading is being done in China, or a Qin Xi if the reading is being done in the US. Ed in another thread mentioned his interset in such ethnic names, and I agree that it seems interesting to record and keep examples of.


Your way is flawed. Are you still going to go on with that? Sure you are...


Please, as I have already encouraged you to do, expand upon your belief.


That way, you are lumping unedited readings from LKL together with edited readings from CO. You don't see a problem with that?


You overlooked the fact that I mentioned that I have a column that says "Live?" (were the readings in the transcript done on live TV, yes or no) to handle that very situation. I guess you didn't quite read what I had posted regarding my current spreadsheet columns for some reason (but yet I'm supposedly incompetent, uh huh).


I am not confusing anything. I gave you the book reference, now you do the work yourself.


I'm not asking you to do work, I'm asking for your constructive suggestions, then you accuse me of asking people to do my work. You keep confusing the two very different things.


Which makes the links irrelevant. You can therefore make all transcripts publically available. Will you?


You have not explained your belief of how that supposedly makes the links irrelevant. You have also not explained why you feel it reasonable for a study to email all people with all updates even people that aren't involved in the study, as well as providing examples where this was done in any other study.


Again, you merely brush off any criticism by referring to it as "belief".


Saying someone is incompetent, without evidence, is not constructive criticism. I have no problem shrugging and brushing off such useless banter.


Please keep focused. We are discussing your flawed analysis here.


I asked you a question and you avoided it. Why are you being so evasive?

But, this coming from someone who says it is worthless to analyze transcripts because of the flawed data, yet freely analyzes his own and puts up articles of people who have analyzed transcripts is quite amusing. Will you address this point ever Claus?

You have also failed to provide evidence that my analysis is flawed. In fact, I haven't even done an analysis yet, so it seems you are very confused here.


So, what are you doing, and for what purpose?

I already have explained such things: We are examining cold reader and medium transcripts (currently of professionals) in a descriptive fashion to see what the data look like as is by recording things that skeptics and believers deem interesting. I seek to have this hopefully cumulative database available, when the study is finished, to both believers, skeptics, and anyone that is interested, as a downloadable Excel file, as a valuable tool for learning about mediums and cold readers.

I also seek to have some notions to questions like, for example, Do mediums use the phrase 'Do you understand that?' often? What is the most frequent letter guessed? Do cold readers and mediums guess ethnic names? How many times have cold readers and mediums guessed 'head or chest areas'? Where are, specifically, all these cold reader and medium transcripts? Do cold readers or mediums do more readings per transcript? Are the letter guesses similar between cold readers and mediums? And many more, regarding this data that has been collected in an observational manner.

I'm obviously not just looking at counts of the words "it" and "and", as some have erroneously suggested. :)

[Gregory]

Originally posted by T'ai Chi

I plan to count "John, George" as 2 names guessed. I'm not sure how to record things like "DJ-sound". I am welcome to hearing your suggestions?

Your project sounds interesting. I don't know why so many people seem to feel the need to attack it.

But I'm not sure how you intend to count guesses. "Fredrick William Shoehorn" is one guess; so is "John." It seems silly to me to count them the same, however, since the first is clearly far less likely than the second.

[CFLarsen]

Originally posted by Gregory
I don't know why so many people seem to feel the need to attack it.

It's not a question of attacking. It's a question of honing the experiment by finding possible flaws. So far, T'ai Chi has been...shall we say: "less than open-minded" about the possibility that he is designing a lousy setup.

It is skepticism at work: It may not be a cuddly-touchy-feely tea party, but it does ferret out the flaws and misconceptions. To some, it is a burden, a nuisance, something that is to be brushed off. To others, it is how we find real knowledge.

[T'ai Chi]

Originally posted by Gregory

Your project sounds interesting. I don't know why so many people seem to feel the need to attack it.

But I'm not sure how you intend to count guesses. "Fredrick William Shoehorn" is one guess; so is "John." It seems silly to me to count them the same, however, since the first is clearly far less likely than the second.

Hi Gregory, who knows why there is so much negativity. Regardless of various detractors' (none of whom have contacted me saying they want to be involved) distraction antics, this study will get, and is getting, done. It has already made some good progress and it has only been roughly 3 days. I'm hoping I can get input and assistance from a variety of people, skeptics and believers.

Your question is a good one! I think I'd currently think of counting such name guesses like 'Fredrick William Shoehorn' as three names guessed, but I am open to other suggestions of course.

I should probably rename my spreadsheet column labelled "NameGuessCount" to be more specific like "AnyNameGuessedCount" per your suggestion, and extend this to include pets too.

[T'ai Chi]

Originally posted by CFLarsen

It's not a question of attacking. It's a question of honing the experiment by finding possible flaws. So far, T'ai Chi has been...shall we say: "less than open-minded" about the possibility that he is designing a lousy setup.


As usual, you are free to offer your comments on how to examine transcripts. You have examined them in the past, yet when I do, suddenly the data is supposedly flawed.

You can explain why you believe this way for us.

[CFLarsen]

Originally posted by T'ai Chi
Hi Gregory, who knows why there is so much negativity.

It's not about "negativity". It's about you designing a fundamentally flawed experiment.

Originally posted by T'ai Chi
Regardless of various detractors' (none of whom have contacted me saying they want to be involved)

Perhaps it is because you are grossly incompetent?

Originally posted by T'ai Chi
distraction antics, this study will get, and is getting, done. It has already made some good progress and it has only been roughly 3 days. I'm hoping I can get input and assistance from a variety of people, skeptics and believers.

But, for some reason, that you will not reveal, it has to be done in secret.

Originally posted by T'ai Chi
Your question is a good one! I think I'd currently think of counting such name guesses like 'Fredrick William Shoehorn' as three names guessed, but I am open to other suggestions of course.

I should probably rename my spreadsheet column labelled "NameGuessCount" to be more specific like "AnyNameGuessedCount" per your suggestion, and extend this to include pets too.

What, in the middle of the experiment, you are going to change the parameters? Sure........!!!

[Paul C. Anagnostopoulos]

T'ai, I still don't understand what you're going to do with these columns after you tally them up. What does it mean for the count of name guesses to be lower or higher?

~~ Paul

[BillHoyt]

Originally posted by Gregory


Your project sounds interesting. I don't know why so many people seem to feel the need to attack it.

But I'm not sure how you intend to count guesses. "Fredrick William Shoehorn" is one guess; so is "John." It seems silly to me to count them the same, however, since the first is clearly far less likely than the second.

There's no probability analysis here, Gregory. Only counting. If you impose any probability assumptions you would be modeling, and on that slippery slope to actually trying to draw an inference from the data. But T'ai says he is simply stamp collecting. Except when he says he isn't and may draw an influence. Even though such an inference would be draw absent a priori model and test decisions and absent any indicator of the mine he had to wade through before coming up with his little nugget. That information, you see, would tell us if the "nugget" is actually significant and we can't have that.

But on the technical merits of your point (as opposed to the demerits of T'ai's), if we compare "John" with "Joshua," and we reckoned probabilty not in the letter arrangement sense with which you are imbuing it, but against the actual distribution of names in the population, we would find that "John" is far more probable than "Joshua."

[T'ai Chi]

Originally posted by CFLarsen

It's not about "negativity". It's about you designing a fundamentally flawed experiment?


There is NO experiment being done. This is an observational study, dude.


But, for some reason, that you will not reveal, it has to be done in secret.


This has already been addressed. If you'd like to be involved, please, as mentioned before, you are welcome to PM me and let me know.

How things can be "secret" when I open a thread to solicit suggestions and talk publicly about what will take place is a mystery to me.

Perhaps you could explain your belief?


What, in the middle of the experiment, you are going to change the parameters? Sure........!!!

1) there is NO experiment.

2) no counting of things in the transcripts has even taken place yet.

3) Number 2) is because this thread's purpose is to obtain constructive suggestions on interesting things to look for in a given transcript before examining the transcripts is actually carried out.

Please let me know if you understand.

[T'ai Chi]

Originally posted by Paul C. Anagnostopoulos
T'ai, I still don't understand what you're going to do with these columns after you tally them up. What does it mean for the count of name guesses to be lower or higher?

~~ Paul

It would mean that in the transcripts we studied, the mediums guessed more, less, or equal amounts of names when compared to the cold readers. :) This might be important to people, believers and skeptics, who are interested in the similarities and differences between mediums and cold readers.

The point is that without actually collecting the data, you have no clue what the reality is.

[T'ai Chi]

Originally posted by BillHoyt

But T'ai says he is simply stamp collecting.


I've said that I'm doing descriptive statistics.


Except when he says he isn't and may draw an influence.


Because you've only imagined that I've said that, that isn't a problem. You seem to ignore the fact that I said something like 'in the data we studied', ie. we can talk about the sample we took- that is what descriptive statistics are.

[Ersby]

Originally posted by BillHoyt


There was a wave of "literary criticism" around in the 60s and 70s. Perhaps it still exists?

What ARE you talking about?

T'ai Chi:

The numbers of times mediums guess a name is unimportant. It's the success rate that matters. Besides, on LKL John Edward guesses names a lot, while JVP doesn't. From that we can assume what, exactly?

I made constructive criticism on page one.

(edited to make clear I was talking to two different people)

[Ed]

A Will 'o the Wisp.

On his thread concering the DAT Model T'ai wrote:

"Asking for opinions is not trolling.

If you've read the article, I'd like to see your criticisms of DAT.

I'm still editing my critique of it."

Never forthcoming.

Any bets on the outcome of this latest piece of work?

[CFLarsen]

Originally posted by Ed
A Will 'o the Wisp.

On his thread concering the DAT Model T'ai wrote:

"Asking for opinions is not trolling.

If you've read the article, I'd like to see your criticisms of DAT.

I'm still editing my critique of it."

Never forthcoming.

Any bets on the outcome of this latest piece of work?

Hehehe....I taught you well, my son... :)

[BillHoyt]

Originally posted by Ersby
I'd have to say all of T'ai Chi's criteria are superficial and meaningless.

Ersby,

I was commenting on this when I posted about that wave of literary criticism. The “critics” were trying to give the appearances of a scientific approach and failing miserably. They literally did as I said: scanned Joyce, Hemingway, etc, into a computer and had it spit out word counts. They then published books filled with these counts.

What was wrong with it? Like T’ai’s “study” here it had the basic trappings of science, but went nowhere. It concluded nothing. It was incapable of concluding anything. It was useless stamp-collecting. At the end of the day, one couldn’t say much about how Joyce wrote or Hemingway or Faulkner. One could only give counts. The efforts were, as you said, superficial and meaningless.

Now T’ai said this earlier defending his study:

" If we make a graph of some counts, for example, and we notice that the ones coded 'TV reading' are higher, for example, then this reveals to us that in the data we studied there is something special with the TV readings that could be explained by sensible things like editing, for example."

I urge everybody to read this sentence a few times and consider: how does a higher count “reveal” anything? These are stastics that are being collected here. They are subject to random fluctuations. Lest us put a collection together as an example:

A : 23 B : 25 C : 30

Okay, now what did we “reveal” about C? That it is higher than A and B. But that is a mathematical statement, not a statistical one. Statistically, without a probability model, we have no idea if C is really higher. Let me now expand the example to say these numbers are from a recent politcal poll, and that the non-tabloid presentation of them ought to have been:

A : 23 +/- 4 B: 25 +/- 4 C: 30 +/- 4

Now it is clearer that these numbers are statistics and that they overlap. What we really know is:

A: [19, 27] B: [21, 29] C: [26, 34]

With these overlapping ranges, we can draw no conclusion that C is statistically higher than A. C is mathematically higher than A, but statistically the same.

So what can we now say about T’ai’s study? Well, we have to conclude one of two things: Either (1) he is ignorant of or ignoring statistical fundamentals in claiming that C being mathematically higher than A “reveals” something or (2) he thinks he can convert this study to an inferential statistics one after the fact. Now number 2 is absolutely wrong, and it would be hard to believe that anybody trained in science or statistics would make such a greenhorn error, I’ll admit. But, T’ai made exactly these claims while discussing triple-blinding a while back. He somehow thinks that it ok to collect data and fit the model to the data after the fact.

Here he confuses exploratory data analysis with inferential statistics. A valid technique would be to collect such data in such an exploratory mode, and then use such data to form an hypothesis and a statistical model. T’ai trips over the next step in the process, though: one must toss out that original data and then go collect new data, and apply inferential techniques to this new data only. The original data can never be used to “reveal... something special with the TV readings.”

[Ed]

Originally posted by CFLarsen


Hehehe....I taught you well, my son... :)

Jesus Christ in Ectoplasm, you have any idea how old that would make you? 35 at least.:D

[Ed]

Originally posted by BillHoyt


Ersby,

I was commenting on this when I posted about that wave of literary criticism. The “critics” were trying to give the appearances of a scientific approach and failing miserably. They literally did as I said: scanned Joyce, Hemingway, etc, into a computer and had it spit out word counts. They then published books filled with these counts.


I recall these analyses. It seems, if memory serves, that they developed algorithms that "wrote" in various styles using the vocabulary and sentence length from their analyses. It was sorta curious since the paras that they generated did sorta look like that of the original authors. This of course was the conceptual granddaddy of the Bible Code fraud.

[Ed]

Originally posted by BillHoyt


So what can we now say about T’ai’s study? Well, we have to conclude one of two things: Either (1) he is ignorant of or ignoring statistical fundamentals in claiming that C being mathematically higher than A “reveals” something or (2) he thinks he can convert this study to an inferential statistics one after the fact. Now number 2 is absolutely wrong, and it would be hard to believe that anybody trained in science or statistics would make such a greenhorn error, I’ll admit. But, T’ai made exactly these claims while discussing triple-blinding a while back. He somehow thinks that it ok to collect data and fit the model to the data after the fact.



or, (3) Nothing of the results will ever appear anywhere except carefully selected snippits to prove some woo-woo point. All couched, naturally, in cautious terms about it being preliminary and so on. Classic woo stuff.

I note that T'ai never explained my hypocracy. That is very, very bad form.

[Ed]

:)

[BillHoyt]

Originally posted by Ed

I recall these analyses. It seems, if memory serves, that they developed algorithms that "wrote" in various styles using the vocabulary and sentence length from their analyses. It was sorta curious since the paras that they generated did sorta look like that of the original authors. This of course was the conceptual granddaddy of the Bible Code fraud.
If you're talking about the same thing I'm thinking about then we're not talking about the same thing at all.


:D

I'll let that stand alone to see if anybody tries to turn it into a sig line.

The analyses I was alluding to were no more complicated than I described. The books contained nothing but the word counts. I think what you're alluding to here was some pretty cool computer work done in the late 70s and early 80s. I have an old issue of Sci Am somewhere here with a good article about it. That work, unlike the "literary criticism" work, actually had meat to it. The computer researchers were trying to see if some sense could be made by carefully randomized letter selection. They took text from various languages, broke it down into conditional probability chains and used those probabilities to randomly generate text that would "look like" the original. IIRC, they found that, with a 6th or 7th letter order analysis, the generated text read pretty well.

In English, for example, given a "t", an "h" is a highly likely next letter. After that, an "e" is fairly likely. "the" "then" "their" "they" "there" If those conditional chains are examined out to 6 or 7 letters (6th or 7th "order"), the randomly produced letters looked very much like English sentences.

[T'ai Chi]

Originally posted by Ersby

T'ai Chi:

The numbers of times mediums guess a name is unimportant. It's the success rate that matters. Besides, on LKL John Edward guesses names a lot, while JVP doesn't. From that we can assume what, exactly?

I made constructive criticism on page one.

(edited to make clear I was talking to two different people)

Ersby, would you suggest counting the hit rate for any given transcript? I'd like to, but it seems rather subjective sometimes. I guess I could assign a skeptical volunteer to count the hit rate for a transcript, and also assign the same transcript to a believer to count the hit rate for a transcript, and see if there is a consensus and maybe take an average of their hit rates if there is not a consensus? What do you think?

[T'ai Chi]

Originally posted by Ed
A Will 'o the Wisp.

On his thread concering the DAT Model T'ai wrote:

"Asking for opinions is not trolling.

If you've read the article, I'd like to see your criticisms of DAT.

I'm still editing my critique of it."

Never forthcoming.

Any bets on the outcome of this latest piece of work?

I gave my thoughts on the DAT model in the thread, hence the whole reason for me saying that if psi doesn't exist, then DAT is just a test for non-randomness.

[T'ai Chi]

Originally posted by BillHoyt

I was commenting on this when I posted about that wave of literary criticism. The “critics” were trying to give the appearances of a scientific approach and failing miserably. They literally did as I said: scanned Joyce, Hemingway, etc, into a computer and had it spit out word counts. They then published books filled with these counts.


Could you give specific references? I'd like to see how what they did is nothing like what I am proposing.


Statistically, without a probability model, we have no idea if C is really higher.


The study is descriptive... not inferential. It describes, summarizes, and displays, the sample.


So what can we now say about T’ai’s study? Well, we have to conclude one of two things: Either (1) he is ignorant of or ignoring statistical fundamentals in claiming that C being mathematically higher than A “reveals” something or (2) he thinks he can convert this study to an inferential statistics one after the fact.


I'm voting for:

(3) Bill has misunderstood several things, including the difference between descriptive and inferential statistics and T'ai Chi's study.


Now number 2 is absolutely wrong, and it would be hard to believe that anybody trained in science or statistics would make such a greenhorn error, I’ll admit.


This from someone who said that skewness is only called skew. Get real.


He somehow thinks that it ok to collect data and fit the model to the data after the fact.


You ARE confused. There is NO model. This is a descriptive study.

[Ed]

All of which begs the question "why bother?"

BTW, where was I hypocritical?

[T'ai Chi]

Originally posted by Ed
All of which begs the question "why bother?"


As mentioned, people talk about cold reader and medium transcripts, but no one seems to have any large scale numerical data on things actually in the transcipts.

Such a thing could allow us to answer questions that could interest believers, skeptics, and anyone interested in mediumship or cold reading.

[Ed]

Originally posted by T'ai Chi


As mentioned, people talk about cold reader and medium transcripts, but no one seems to have any large scale numerical data on things actually in the transcipts.

Such a thing could allow us to answer questions that could interest believers, skeptics, and anyone interested in mediumship or cold reading. [/B]

What sort of questions? Suppose you were writing a grant proposal, what would the reason for being for this study be? Certainly more than "nice to know" I hope. Why should one care?

Where was I hypocritical?

[T'ai Chi]

I also thought of one area where science is often done where there is often secrecy about current studies and even studies reaching far back, as a counter-example: intelligence analysis.

[Garrette]

T'ai Chi

I also thought of one area where science is often done where there is often secrecy about current studies and even studies reaching far back, as a counter-example: intelligence analysis.

Excuse me?

In what context, please? If you mean military, you are mistaken. {And, yes, I have the credentials to back this up.} Analytical methods will incorporate analytical tools which may have a scientific origin, but the analysis is not in itself scientific, anymore than a criminal investigation is scientific.

Possession of scientific attributes does not by itself gain an endeavor scientific status.

If you mean some other context for intelligence analysis, please explain. Another application that comes to mind for me is Corporate Intelligence which models itself largely on Military Intelligence but is further removed from being science by the fact that it is even more restricted by criminal statutes.

Which leaves, I think, organizations like NSA and the CIA. Again, their analysis uses scientific tools but is not science. Some decryption efforts are highly mathematical, but they are not put forth as experiments to support a hypothesis; the intelligence, which is never (properly, at least) put forth as fact is used for policy development, not for science. The same is true of node analysis when working with communications trees.

You're off base here, T'ai.

[CFLarsen]

Originally posted by Garrette
You're off base here, T'ai.

No. He is desperate to come up with an explanation why he wants to keep it secret. To some.

He wants to cut off information, but only to some. And that's where he shoots himself in the foot: Not only does he have to explain why he wants to keep it secret, he also has to explain why those people can know.

Are those people under oath? Have they made a pledge not to reveal anything? How is T'ai Chi going to secure that no one among these selected few keep their mouth shut? Is that part of the protocol as well?

Of course he can't stop anyone from spilling the beans. Of course there are no reasons to keep it secret. The only reasons T'ai Chi wants to keep it secret I can think of are:


He is just doing it to tease.
He wants to boost the hype.
He wants to make it impossible to see how he got his results.


A combination of these are, of course, a possibility.

[T'ai Chi]

Originally posted by Ed

What sort of questions?


What would you like to know? :)

In the data we collected we could investigate things like: Are cold reader readings similar to medium readings? What letters do mediums tend to guess? How often to mediums use the 'head or chest area' as a region involved with the death? Or how often do mediums say 'Do you understand that?' or 'Take that home with you' type of responses?


Where was I hypocritical?

Ed, I said (underline added) "Considering I've addressed your point and found it unconvincing and hypocritical, your above points are rather vacuous." where your point being that analyzing transcripts is a waste of time and that the results mean nothing. It is hypocritical because skeptical analysis of medium transcripts has been carried out (by the same people who are trying to give me that line that analyzing transcripts is worthless. I, myself, think that analyzing transcripts doesn't show a whole lot, but, it is interesting, worth exploring, and can investigate some worthwhile questions at least, in my opinion).

[T'ai Chi]

Originally posted by Garrette

In what context, please?


In the context of science being carried out by government agencies, military, etc., where classification is necessary.


Analytical methods will incorporate analytical tools which may have a scientific origin, but the analysis is not in itself scientific, anymore than a criminal investigation is scientific.


A good deal of criminal investigation techniques are very scientific.


Possession of scientific attributes does not by itself gain an endeavor scientific status.


I agree with that, but that doesn't negate my point.

[T'ai Chi]

Originally posted by CFLarsen

He is desperate to come up with an explanation why he wants to keep it secret. To some.


Are your recent dabblings in mind-reading supposed to impress anyone?

I was merely giving a counter-example to the common 'science is not done in secrecy' line.

As stated, your mail and email box must be overflowing with all the updates you continually get from researchers all over the world, Claus. You have yet to name even one study where updates were given to anybody during every stage of the study, such as you are holding my simple counting study to. Will you ever?


He wants to cut off information, but only to some.


LOL! I've freely invited those intersted to PM me if they want to receive updates when the project is in full swing. Those who are "cut off",as you say, are by their own choice.


Are those people under oath? Have they made a pledge not to reveal anything? How is T'ai Chi going to secure that no one among these selected few keep their mouth shut?


Huh?


He wants to make it impossible to see how he got his results.


Considering I started this thread to solicit people for their suggestions on how to look and get results, I find your belief highly improbable.

[CFLarsen]

T'ai Chi,

I would like to receive updates. My email is webmaster@skepticreport.com

When can I expect the first update?

[Garrette]

T'ai Chi:

In the context of science being carried out by government agencies, military, etc., where classification is necessary.


Okay, so you're not talking about intelligence analysis at all, then; you're talking about experiments. The old CIA Cold War psi stuff, maybe?

You'll note that none of the rumored claims of those things were accepted until details were revealed, at which time it was discovered there was no there there.

T'ai Chi:

A good deal of criminal investigation techniques are very scientific.


Exactly my point. Some of the tools used in support of an investigation (or intelligence analysis) may be scientific--and I assume you're thinking along the lines of forensics--but do not make the investigation itself scientific. Were it so, then juries would be unnecessary.

More to the point, these are, as you say, scientific techniques, not scientific experiments.

Even more to the point, the data collected which is, I think, analogous to what you're proposing, such as the FBI's statistics on what times are committed most, where, by whom, etc. (I'm simplifying here) are not kept secret.

Nor are the scientific portions of criminal investigations. They must, in fact, be made available to the defense which can then present them in court.

No secrecy.

T'ai Chi:

I agree with that, but that doesn't negate my point.

Yes, it does.

[T'ai Chi]

Originally posted by CFLarsen
T'ai Chi,

I would like to receive updates. My email is webmaster@skepticreport.com

When can I expect the first update?

You are welcome to PM me then as instructed. How many times must I repeat myself, Claus? :)

[CFLarsen]

Originally posted by T'ai Chi
You are welcome to PM me then as instructed. How many times must I repeat myself, Claus? :)

I don't use PM. Is it a problem to use email?

[T'ai Chi]

Originally posted by Garrette

Okay, so you're not talking about intelligence analysis at all, then; you're talking about experiments. The old CIA Cold War psi stuff, maybe?


No, I'm talking about science being carried out by government agencies, military, etc., where classification is necessary.


Some of the tools used in support of an investigation (or intelligence analysis) may be scientific--and I assume you're thinking along the lines of forensics--but do not make the investigation itself scientific. Were it so, then juries would be unnecessary.


Juries are needed because that is how the juducial system works. The presence of juries doesn't mean science has not taken place. It might surprise you to know there are even juries (peer review) in science.


More to the point, these are, as you say, scientific techniques, not scientific experiments.


Ok. I guess we are debating whether or not an area that often employs 'scientific techniques' can be called 'science'. In forensics there are experiments though.


No secrecy.


Sure, in some cases there isn't, and in some there is. I'm not saying there is never any secrecy, only that there is often secrecy due to classification and the need of national security, sensisitive information, and etc.

[T'ai Chi]

Originally posted by CFLarsen

I don't use PM. Is it a problem to use email?

"as instructed". What are you failing to understand? Beacuse I have explained myself many times on this, I can only assume you are attempting to make a point here. You have made your point: you can't follow directions.

I'm not too interested in troublemakers. Either PM me if you are interested, as instructed, or don't. :)

[CFLarsen]

T'ai Chi,

I have PM'd you. Please keep me in the loop by emailing me.

[Ed]

Originally posted by T'ai Chi



Ed, I said (underline added) "Considering I've addressed your point and found it unconvincing and hypocritical, your above points are rather vacuous." where your point being that analyzing transcripts is a waste of time and that the results mean nothing. It is hypocritical because skeptical analysis of medium transcripts has been carried out (by the same people who are trying to give me that line that analyzing transcripts is worthless. I, myself, think that analyzing transcripts doesn't show a whole lot, but, it is interesting, worth exploring, and can investigate some worthwhile questions at least, in my opinion). [/B]

Got it. You either misstated or used the word incorrectly.

[T'ai Chi]

Originally posted by CFLarsen
T'ai Chi,

I have PM'd you. Please keep me in the loop by emailing me.

As of now, your PM you said you sent is not in my PM inbox. Perhaps it takes a while to go through though.

I'm sure you can back up your claim that you actually sent a PM though in the case it does not go through..

[T'ai Chi]

Originally posted by Ed

Got it. You either misstated or used the word incorrectly.

Neither.

As stated, I find the point, that me analyzing transcripts is worthless, but so called skeptics can analyze transcripts, to be hypocritical.

[CFLarsen]

Originally posted by T'ai Chi
As of now, your PM you said you sent is not in my PM inbox. Perhaps it takes a while to go through though.

I'm sure you can back up your claim that you actually sent a PM though in the case it does not go through..

Of course.

Has it gone through yet?

[Ed]

Originally posted by T'ai Chi


Neither.

As stated, I find the point, that me analyzing transcripts is worthless, but so called skeptics can analyze transcripts, to be hypocritical.

And I am hypocritical how?

[BillHoyt]

Originally posted by CFLarsen
The only reasons T'ai Chi wants to keep it secret I can think of are:


He is just doing it to tease.
He wants to boost the hype.
He wants to make it impossible to see how he got his results.


A combination of these are, of course, a possibility.
There is a fourth possibility, hinted at by his slip into inferential statistics. Recall also his assertion about triple-blind studies: that one can choose significance levels and models after seeing the data. Put them all together and one has the strange speculation that he will collect this data in secret, mine it for "significant" results and report those significant results.

In this case, it is not simply a matter of making it impossible to see how he got his results, but making it impossible to determine whether the results are actually significant. If he does 20 analyses at the .05 level, we expect one to show significance by random fluctuation alone. That is, it ain't really significant. If he hides the other 19, of course, we don't know that is what happened.

This statement of his is a statement requiring inferential statistics. It cannot be supported with descriptive statistics:

" If we make a graph of some counts, for example, and we notice that the ones coded 'TV reading' are higher, for example, then this reveals to us that in the data we studied there is something special with the TV readings that could be explained by sensible things like editing, for example."

[CFLarsen]

Originally posted by BillHoyt
If he hides the other 19, of course, we don't know that is what happened.

Exactly: We are not allowed to see just how many transcripts he goes through. The Filedrawer effect.

A thoroughly dishonest procedure.

[Ed]

Originally posted by BillHoyt

In this case, it is not simply a matter of making it impossible to see how he got his results, but making it impossible to determine whether the results are actually significant. If he does 20 analyses at the .05 level, we expect one to show significance by random fluctuation alone. That is, it ain't really significant. If he hides the other 19, of course, we don't know that is what happened.


Yes, this calls to mind that dead fraud, Targ, as well as a couple of woo "prayer effectiveness" studies that I have seen.

The point is that by doing this one is simply lieing. It is not as though any of these events are "slightly" significant. BTW, Bill, would you agree with the rule of thumb of adjusting the significance level liniarly by the number of tests? So for 10 tests p starting out at .05 would be .005?

[Garrette]

T’ai Chi

No, I'm talking about science being carried out by government agencies, military, etc., where classification is necessary.

Okay. But this is an important change from science in military intelligence or intelligence analysis. If it was just misspeaking when you first brought it up, no problem. Was it?

But that leads to the question of what reason you are giving for keeping your information secret. Are there reasons of national security?

And I refer to my earlier post that even if classified experiments are done, they are not accepted by the scientific community until, well, until the scientific community has a chance to see them.


T’ai Chi

Juries are needed because that is how the juducial system works. The presence of juries doesn't mean science has not taken place. It might surprise you to know there are even juries (peer review) in science.

You’ll need to define science. If, in a forensics lab, I apply a drop of Chemical X to Evidence Y and by Reaction Z conclude that Suspect A was indeed at Crime Scene B, then I suppose it’s science, but it’s not science the way you are attempting to imply here. This is just an application of proven/accepted techniques. It is not an attempt to prove a new theory or demonstrate some fact heretofore unknown to the scientific community. They are not analogous.

And, no, it does not surprise me at all that peer review (juries) is conducted in science; I know this. It might surprise you to learn that this is not remotely analogous to a trial jury.

I can theoretically serve jury duty in court, but no scientific journal in the world would or should accept me as part of a scientific peer review process. The qualifications for the first are citizenship; the qualifications for the second are relevant technical knowledge.

Funny thing about it is that I will likely never serve on either type of jury. I’ve been called for jury duty a few times, but always rejected because I know too much and can think. For the peer reviews, I’d be rejected because, although I can think, I don’t know enough.



T’ai Chi

Ok. I guess we are debating whether or not an area that often employs 'scientific techniques' can be called 'science'. In forensics there are experiments though.

Sure, in some cases there isn't, and in some there is. I'm not saying there is never any secrecy, only that there is often secrecy due to classification and the need of national security, sensisitive information, and etc.

Again, the forensic experiments are simply applications of known principles, not searches for new ones.

I know you’re not saying there is never any secrecy; this came up because your claim was more toward the opposite end: that secret experiments exist and are accepted.

Again, what is your reason for keeping it secret? National security? Sensitive information?

[Ed]

Originally posted by Garrette


Again, what is your reason for keeping it secret? National security? Sensitive information?

Fear of derision? Ability to cherry-pick results?

[CFLarsen]

Originally posted by Garrette
Again, what is your reason for keeping it secret? National security? Sensitive information?

Originally posted by Ed
Fear of derision? Ability to cherry-pick results?

No reason. You see, the data is all available, so why would T'ai Chi need to keep it secret?

....waaaaaaait. The data is not all available!

Silly us! http://www.skepticreport.com/resources/smilies/slap.gif

[BillHoyt]

Originally posted by Ed
The point is that by doing this one is simply lieing. It is not as though any of these events are "slightly" significant.
Exactly so.

BTW, Bill, would you agree with the rule of thumb of adjusting the significance level liniarly by the number of tests? So for 10 tests p starting out at .05 would be .005?
A Bonferroni adjustment should be used in this case, but Bonferroni has several big problems.¹. The corrected alpha for Bonferroni is: 1-(1-alpha)^1/n The most significant problem is that it makes a Type I / Type II tradeoff. While it assures the overall Type I error (incorrectly rejecting the null hypothesis) remains at the original alpha level, it does so by increasing the Type II error (incorrectly accepting the null hypothesis).

Of course, in a previous exchange with T'ai about Bonferroni, he somehow had the idea that Bonferroni creates a new alpha level. It doesn't. It creates a correction that attempts to preserve the original alpha level under mulitple hypotheses.

______
^{1 (http://bmj.bmjjournals.com/cgi/content/full/316/7139/1236?view=full&pmid=9553006)}Perneger TV. What's wrong with Bonferroni adjustments. BMJ 1998;316:1236-1238.

[Ed]

Originally posted by BillHoyt

Exactly so.


A Bonferroni adjustment should be used in this case, but Bonferroni has several big problems.¹. The corrected alpha for Bonferroni is: 1-(1-alpha)^1/n The most significant problem is that it makes a Type I / Type II tradeoff. While it assures the overall Type I error (incorrectly rejecting the null hypothesis) remains at the original alpha level, it does so by increasing the Type II error (incorrectly accepting the null hypothesis).

Of course, in a previous exchange with T'ai about Bonferroni, he somehow had the idea that Bonferroni creates a new alpha level. It doesn't. It creates a correction that attempts to preserve the original alpha level under mulitple hypotheses.

______
^{1 (http://bmj.bmjjournals.com/cgi/content/full/316/7139/1236?view=full&pmid=9553006)}Perneger TV. What's wrong with Bonferroni adjustments. BMJ 1998;316:1236-1238.

It appears that for all intents and purposesthe quick and dirty division method is pretty darn close though a bit more rigid.

If this were applied, or if sceptics were versed in it's use I suspect that a passel of questionable research would be dismissed.

Which brings to mind a pet peeve of mine. The people conducting this research should know that one should, nay, must do this yet they often do not. I personally find it damning that this occurs, not just for the credibility of the researchers, but for anyone who is aware of the problems and remains uncritical.

[Ed]

A follow up re. data mining.

What is wrong with this picture? Go to the table on page 2.

http://www.godandscience.org/apologetics/smj1.html

[BillHoyt]

Originally posted by Ed
A follow up re. data mining.

What is wrong with this picture? Go to the table on page 2.

http://www.godandscience.org/apologetics/smj1.html

First thing wrong is I can't get to the link. Is it correct?

[Ed]

Originally posted by BillHoyt


First thing wrong is I can't get to the link. Is it correct?

Well, I was on their site, perhaps they have limited bandwidth?

Seriously, it works for me. The table is smj2

It is a PDF file, btw.


NOTE: Would non-US posters (actually US posters too) take a look at the thread I started in the main fourm entitled Querry or something like that. I am looking for something.

[T'ai Chi]

Originally posted by CFLarsen

Of course.

Has it gone through yet?

No, it is not in my PM inbox. Please try again.

[T'ai Chi]

Originally posted by Ed


And I am hypocritical how?

I don't believe I said you were Ed (I'm open to you findin evidence I did), just the point that skeptics make that analyzing transcripts is worthless yet they analyze transcripts.

[T'ai Chi]

Originally posted by BillHoyt

A Bonferroni adjustment should be used in this case, but Bonferroni has several big problems.¹. The corrected alpha for Bonferroni is: 1-(1-alpha)^1/n The most significant problem is that it makes a Type I / Type II tradeoff. While it assures the overall Type I error (incorrectly rejecting the null hypothesis) remains at the original alpha level, it does so by increasing the Type II error (incorrectly accepting the null hypothesis).


You don't accept a null hypothesis, you fail to reject them.


Of course, in a previous exchange with T'ai about Bonferroni, he somehow had the idea that Bonferroni creates a new alpha level. It doesn't. It creates a correction that attempts to preserve the original alpha level under mulitple hypotheses.


Anyway..

[Ed]

Originally posted by T'ai Chi


You might not have, sure.

[/b]

Then would you agree that skeptics analyzing transcripts for cold reading is questionable?

Since I'm not doing inference, I'm merely describing the data that is out there, it isn't a big deal. As mentioned, we can create a column in the spreadsheet that records if the transcript is from a TV reading, or a live reading, or from a book, etc. If we make a graph of some counts, for example, and we notice that the ones coded 'TV reading' are higher, for example, then this reveals to us that in the data we studied there is something special with the TV readings that could be explained by sensible things like editing, for example. We won't know until we get the actual data. Again, keep in mind I've pretty much decided to not look at subjective things like hits or misses.

[/b]

You have yet to really convince that it is unworthy of study and that edited = flawed. According to you, one can never analyze any transcript, anything on TV, anything in book form, anything in any type of media (except completely live recordings/movies) which is surely a worthless way of looking at things because nothing ever gets done. I disagree with it 100%, for reasons I mentioned above.



Considering I've addressed your point and found it unconvincing and hypocritical, your above points are rather vacuous. [/B]

Last paragraph, T'ai.

[Ed]

Originally posted by T'ai Chi


I don't believe I said you were Ed (I'm open to you findin evidence I did), just the point that skeptics make that analyzing transcripts is worthless yet they analyze transcripts.

And I have said that edited transcripta are useless.

[T'ai Chi]

Originally posted by Garrette

But that leads to the question of what reason you are giving for keeping your information secret. Are there reasons of national security?


No secrecy, and of course its not a matter of national security. I merely brought up some science done by government agencies that is classified and a matter of national security as being examples of science done in secret.

Your email and mail box must be overflowing with updates from studies conducted from all over the world. :)

Garrete, there is no secrecy with my propose study. If you want to be one of the participants, PM me. If you have constructive suggestions on how to examine transcripts, let this thread know. The full Excel spreadsheet will be made available after the study is done.

Please name one study in which updates, at any stage of the study, are made available to all who request them, for whatever reason. Just one. :)


And I refer to my earlier post that even if classified experiments are done, they are not accepted by the scientific community until, well, until the scientific community has a chance to see them.


Oh, within their community they'll be seen, just not, at the moment at least, by the general public.


You’ll need to define science.


:rolleyes: I see, you can talk about science without defining it, yet when I talk about it, I need to define it.


Again, the forensic experiments are simply applications of known principles, not searches for new ones.


You don't have to continually discover new laws to be doing science. Why do you think you do??


Again, what is your reason for keeping it secret? National security? Sensitive information?

Again, your email and mail box must be overflowing with updates from studies conducted from all over the world. :)

Garrete, there is no secrecy with my proposed study. If you want to be one of the participants, PM me. If you have constructive suggestions on how to examine transcripts, let this thread know. The full Excel spreadsheet will be made available after the study is done.

Please name one study in which updates, at any stage of the study, are made available to all who request them, for whatever reason. Just one. :)

[T'ai Chi]

Originally posted by CFLarsen

Exactly: We are not allowed to see just how many transcripts he goes through. The Filedrawer effect.


The Excel spreadsheet will contain each and every transcript, with exact reference, that was examined.

In any case, that's not what a filedrawer effect is. A filedrawer effect is when one does a meta-analysis but has a hypothetical filedrawer containing non-significant studies that the researcher didn't include in the meta-analysis.

[T'ai Chi]

Originally posted by BillHoyt

There is a fourth possibility, hinted at by his slip into inferential statistics.


As already mentioned, there was no slip. This is a descriptive study, and I said something like 'in the data we studied'.


Recall also his assertion about triple-blind studies: that one can choose significance levels and models after seeing the data.


Please "Recall" for us exactly where I supposedly said what you attributed to me above. You are 'fool of hot air'.


Put them all together and one has the strange speculation that he will collect this data in secret, mine it for "significant" results and report those significant results.


Is your psychic intuition supposed to impress anyone? I don't plan to do any of that, as already stated.


If he does 20 analyses at the .05 level, we expect one to show significance by random fluctuation alone. That is, it ain't really significant. If he hides the other 19, of course, we don't know that is what happened.


What in the world are you talking about? I've already stated, but I'll repeat and hope maybe this time you'll listen, that NO INFERENCE WILL BE DONE. What don't you understand here? :)


This statement of his is a statement requiring inferential statistics. It cannot be supported with descriptive statistics:

" If we make a graph of some counts, for example, and we notice that the ones coded 'TV reading' are higher, for example, then this reveals to us that in the data we studied there is something special with the TV readings that could be explained by sensible things like editing, for example."

Bill, did you overlook: "in the data we studied"?

[T'ai Chi]

Originally posted by Ed


Last paragraph, T'ai.

I've already addressed this. Underlines added:

"Considering I've addressed your point and found it unconvincing and hypocritical, your above points are rather vacuous."

The point you are making is hypocritical as I've mentioned. I didn't say you were a hypocrite, since I am aware that you think analzying all edited transcripts is worthless. My point is that skeptics have analyzed transcripts and some of these same people are now saying that me proposing to analyze transcripts is useless because of the flawed data, etc.

[Ed]

Originally posted by T'ai Chi


I've already addressed this. Underlines added:

"Considering I've addressed your point and found it unconvincing and hypocritical, your above points are rather vacuous."

The point you are making is hypocritical as I've mentioned. I didn't say you were a hypocrite, since I am aware that you think analzying all edited transcripts is worthless. My point is that skeptics have analyzed transcripts and some of these same people are now saying that me proposing to analyze transcripts is useless because of the flawed data, etc.

Makes no sense. How can a point I make be hipocritical while I am not? I couldn't give a flying one about what "sceptics" do, the only person I am responsible for is me.

OK, let's drop it.

[Garrette]

All quotations originally posted by T’ai Chi:

No secrecy, and of course its not a matter of national security.
I think we may have a misunderstanding here on what you will make available via e-mails after requests through your pm. More on this later.

I merely brought up some science done by government agencies that is classified and a matter of national security as being examples of science done in secret.
Actually, you merely said it’s being done. You brought up no science.


Your email and mail box must be overflowing with updates from studies conducted from all over the world.
Correct me if I’m wrong, but studies don’t get updated continuously. An experiment/study is conducted. Findings are published (or otherwise made available). The findings, in my usage here, includes a description of protocols, listing of all data collected/analyzed, and all relevant information. If something was considered, it is listed, even if discarded. This avoids the filedrawer problem, as I understand it.

Garrete, there is no secrecy with my propose study. If you want to be one of the participants, PM me.
Now is the time to address this, I think. I don’t think you are intentionally trying to be secretive, but it appears that you are being so anyway (or that you will be so according to how you will set this up.
You think that since you will make your spreadsheet available to those who ask properly you are being open.
But the objection isn’t about the spreadsheet itself; it’s about what was reviewed to create the spreadsheet.
What spreadsheets did you gather? How did you find them? How did you choose which to use and which to discard? How did you decide your column headings? How did you decide which bits from the transcripts go in which column?
It is the background stuff that seems will be kept secret. Am I mistaken?

If you have constructive suggestions on how to examine transcripts, let this thread know.
I’m not really qualified to do so, but if something really grabs my interest, I may jump in, thanks.
The full Excel spreadsheet will be made available after the study is done.
This is the point. See above.

Please name one study in which updates, at any stage of the study, are made available to all who request them, for whatever reason. Just one.

Now you’re being obtuse. You are the one who has decided only to disseminate via e-mails. That is your own restriction; do not complain to others here if it’s an onerous burden. You can easily overcome it by posting it all via the most public means available to you. This board is an excellent option.
Studies published in journals are “updated” when new studies substantiate, modify, or repudiate their findings. The original studies, if legitimate, include all the background information when originally published.
What is not done, is to put a half-completed study in a journal with comments like:
“No conclusions can be drawn from the following because I’m not done. Also, I don’t think it is pertinent to list my protocols, my hypothesis, nor my models. I hope you find this helpful.”

Oh, within their community they'll be seen, just not, at the moment at least, by the general public.
There’s a difference between publishing something in a medium which is open to the general public but generally not accessed by it and publishing something in a manner which precludes access at all.
There is also a difference between publishing a spreadsheet of numbers and publishing the methods by which that spreadsheet was developed.

I see, you can talk about science without defining it, yet when I talk about it, I need to define it.
When a loose definition suffices for the discussion, I have no problem with it; when distinctions become relevant to the topic, then specificity is called for.
That being said, you’re correct; I should have been more explicit in my usage. My apologies.
So here goes: It seems to me that you are using “science” in a broad sense, meaning something along the lines of “that which uses either scientific apparatus or scientific technique or the application of technical knowledge to answer a question of fact.”
If that (or something similar) is your definition, then I agree that criminal investigations and intelligence analysis are scientific. But if this is the case, it’s such a loose usage that it adds nothing of value to your argument.
For it to be of any use to you, you need to define science more precisely (and, imo, more accurately) along the lines of: “that which uses accepted methodologies to evaluate facts and observations in order to substantiate or repudiate a given hypothesis.”
If your definition is anything along these lines, then your contention about criminal investigations and intelligence analysis being scientific is incorrect.
You used “science” in your claim about intelligence analysis. It was, in fact, the central point. Given that, it is not unreasonable for me to ask you to be specific about it.
If there is a contention of mine which rests on the definition of “science,” then you would be correct to ask me about my usage. Is there such a contention?

You don't have to continually discover new laws to be doing science. Why do you think you do??
See above. This is the key point about this, isn’t it? How were you using the word?

Again, your email and mail box must be overflowing with updates from studies conducted from all over the world.

See above answers and comments.


Garrete, there is no secrecy with my proposed study. If you want to be one of the participants, PM me.

Also see above.

The restrictions on distribution are of your own making. If I request “updates,” will I also receive all the background information, too?


Please name one study in which updates, at any stage of the study, are made available to all who request them, for whatever reason. Just one.

Let’s check out the obverse of this question to see what it looks like then, shall we?

Please name one accepted study which does not release all the relevant data (or, to use my term, background information)?

Your bit about my e-mail being full is a straw man, T’ai. You have chosen e-mail as your method of distributing your spreadsheet, not I or anyone else on this board.

If you chose instead to publish it on skepticreport, that would be fine by us. If Ed publishes something in one of his professional journals, that is updating the community and making data available.

I wager, too, that if Ed published a study, he would include as a matter of course all the background information, and if further studies indicated that his original study were incorrect he would publish that, too, along with all background information again.

[CFLarsen]

Originally posted by T'ai Chi
No, it is not in my PM inbox. Please try again.

I just PM'd you - again.

Originally posted by T'ai Chi
The Excel spreadsheet will contain each and every transcript, with exact reference, that was examined.

In any case, that's not what a filedrawer effect is. A filedrawer effect is when one does a meta-analysis but has a hypothetical filedrawer containing non-significant studies that the researcher didn't include in the meta-analysis.

Wrong, T'ai Chi. Filedrawer effect is "when only positive results tend to get reported and negative ones are left in the filing cabinet."
Source (http://www.csicop.org/sb/2002-12/reality-check.html)

It doesn't have to be a meta-analysis. It simply means: Leave out the results that you don't like.

Originally posted by Garrette
I think we may have a misunderstanding here on what you will make available via e-mails after requests through your pm. More on this later.

There is no misunderstanding: T'ai Chi was very clear that he will:

Originally posted by T'ai Chi
I will PM those who have PM'd me showing their interest. Once the study actually gets in full swing, communication will be done by PM's or email, not in threads here. If you'd like to be in the loop and offer constructive criticism, just PM me and let me know.

Originally posted by Garrette
If you chose instead to publish it on skepticreport, that would be fine by us.

It would be fine by me, too.

[T'ai Chi]

Originally posted by CFLarsen

I just PM'd you - again.


I just got your PM's. I think it was because I had you on my Ignore list that your first one didn't go through? Anyway, you'll be in the group that gets notified of any news.


Wrong, T'ai Chi. Filedrawer effect is "when only positive results tend to get reported and negative ones are left in the filing cabinet."
Source (http://www.csicop.org/sb/2002-12/reality-check.html)

It doesn't have to be a meta-analysis. It simply means: Leave out the results that you don't like.


The author is talking about meta-analysis throughout that entire article.


It would be fine by me, too.

I have no idea why people are even bringing up SkepticalReport. Maybe you could explain why I should be interested in the slightest?

[CFLarsen]

Originally posted by T'ai Chi
I just got your PM's. I think it was because I had you on my Ignore list that your first one didn't go through? Anyway, you'll be in the group that gets notified of any news.

But how can you have me on ignore, if you see and respond to my posts?? You did not have me on ignore. Or, did you put me on ignore, the moment after you asked me to PM you, and a few moments later took me off ignore? That is the only way your explanation is valid.

Stop playing these juvenile games, T'ai Chi. You can email me at webmaster@skepticreport.com when notifications/updates/etc. occur.

Originally posted by T'ai Chi
The author is talking about meta-analysis throughout that entire article.

No, he specifically talks about the Filedrawer effect as being general.

Originally posted by T'ai Chi
I have no idea why people are even bringing up SkepticalReport. Maybe you could explain why I should be interested in the slightest?

It's "SkepticReport". Why wouldn't you? It provides a forum where you can publish your research. Where else would you publish it?

[Garrette]

CFLarsen:

There is no misunderstanding: T'ai Chi was very clear that he will:
{snip quotation of T'ai Chi}

You are referring to the method of disclosure. My comment was in reference to what is being disclosed.


T'ai Chi

I have no idea why people are even bringing up SkepticalReport. Maybe you could explain why I should be interested in the slightest?

I only brought it up as an example, on the assumption that it will not make it into some other accessible journal.

Another alternative is to post it on it's own thread here, complete with all the background and supporting information.

[Ersby]

Ages ago T'ai Chi asked if some measure of hit rate should be made. I'd say yes, but of course, there are hits and there are hits. I sat down and started to list the things I'd look for in a reading, and I think that T'ai Chi is seriously underestimating the amount of work involved if he's going to come up with anything that actually means something.

Here's the categories I think should be included...


Basic Hit rate (in which a hit is when the sitter responds to something in a positive way. This includes everything, even repeating back information and “Do you understand?”)

No. of misses

No. of unacknowledged statements (a statement made by the medium which is never addressed by the sitter)

No. of “Do you understand”

No. of repeating back information

No. of common sense conclusions (whereby if the medium is told they sold a house, he guesses the house needed a little work and gets a hit. This includes guesses about jewellery and watches for older deceased, and guesses about toys for younger energies)

No. of guesses made repeatedly and adjusted by medium

No. of guesses stretched by sitter to fit something

Underlying hit rate (Basic hit rate minus the DYU count, the repeating count, the common sense count, the adjusted guess count and the stretched hit count)


Names:

No. of name guesses

No. of hits

No. of misses

No. of unacknowledged

No. of full-name guesses referred implicitly to one person (was your father called Michael?)

No. of exact hits for those guesses (ie, the father was called Michael)

No. of stretched hits for those guesses (ie, the son was called Michael)

No. of misses for those guesses

No. of unacknowledged full-name single person guesses

No. of full-name guesses to a range of people (ie, is your grandfather called Michael)

No. of exact hits for those guesses

No. of stretched hits for those guesses

No. of misses for those guesses

No. of unacknowledged full-name range guesses

No. of full-name guesses linked to another guess (ie, I’m getting a problem with the lungs, and there’s a Michael connection too)

No. of exact hits for those guesses (ie, Michael had lung problems)

No. of stretched hits for those guesses (both Michael and lung problems mean something, but are not directly linked)

No. of misses for those guesses (miss on Michael)

No. of unacknowledged full-name guesses linked to another guess

No. of full name guesses with no context (i.e., Who’s Michael?)

No. of exact hits for those guesses (i.e., Michael was a person referred to in the reading before the guess)

No. of stretched hits for those guesses (i.e., Michael is someone not mentioned before)

No. of misses for those guesses

No. of unacknowledged full-name no-context guesses


This would then be followed by the same categories for “Choice of full names” (ie, John, Joe) and “Initials” (Peter, P name).

Then it would further followed up by a measure of the probability of getting a hit particular name, taking into account the range of possible targets.

As you see, the names alone constitute an enormous amount of work if they’re to be done properly. Then of course you must understand that if the medium asks about “Michael” but the sitter replies “Mark” and it is accepted by the medium, then that goes into the “Initial” category. Although the medium guessed only one name, it was understood by both parties that any male M name would be accepted.

And that's only the beginning of what should be examined!

[Ersby]

I said before this was a subject I'd given a lot of thought to, so I thought I'd dig out an old analysis I made of a very simple reading (by JE on LKL) to see what people made of it. I just need someone to do the sums at the end, if anyone's willing.

CALLER: I'm calling about my mother.

EDWARD: OK, stop right there. Your first name?

CALLER: Linda.

KING: Just give us your name and who you're asking for.

CALLER: My mother.

EDWARD: OK, Linda, the first thing I want talk about is, I know you're looking for your mom but I'm getting an older male who's also there on the other side. I feel like this is somebody who would be above you, which means it's like a father figure, or an uncle, and he passes from either lung cancer or emphysema, tuberculosis; it's all problems in the chest area. OK, that's the first thing. And I feel like there's a J or a G-sounding name attached to this.

CALLER: That's my mother.

EDWARD: She's got a very dominant personality.

CALLER: That's my mother. Her first name starts with G and she had emphysema.

EDWARD: Hold on. Does the month of August have a meaning for her, or the 8th of a month?

CALLER: Not that I know of.

EDWARD: OK, I want you to write this down, because she's telling me to say "eight" then. I have to tell you that this is coming through so strong there's a male, it's got a very dominant energy, but this is how I'm interpreting it, and she's telling me to talk about "eight." "Eight" to me would indicate that the month of August has a meaning, or that the eighth of a month has a meaning. She's telling me that there's a father-figure that's there, so I don't know if your father's passed but there's a father-type figure...

CALLER: No, my father -- I just spoke to him on my son's phone and he wanted me to ask...

EDWARD: Wait a second.

KING: He's nodding, yeah, your mother was tough.

EDWARD: There's a father figure...

KING: I can see your father, yeah.

EDWARD: ... who's with her, from what she's showing me.

CALLER: Her father?

EDWARD: It's not her father. It's connected to you. So I don't know if there's a father-in-law for you who's passed, but there's a father-figure who's there. It's a male figure who's there.

KING: But the important thing is, how is she doing?

EDWARD: Your mom is fine and I think it's important that you know that she was around -- somebody missed seeing her from what she's showing me, and she's telling me to let you know that.


The name Linda was most popular during the 1950’s and 1960’s. From this, I’m going to assume (yes, ASSUME!) the sitter is between 40 and 50 years old. We’ll say 45. Plus, JE immediately knows her mother has passed, which also indicates a late-to middle aged women. Plus, of course he can hear her voice.

Older male: This means the generation above. The chances of a 65 year old man being dead is 0.18 (I worked this out from a pdf from the National (US) Vital Statistics Report which showed that of 100,000 people alive at birth, 82,000 people are still alive at 66) Older male could be grandfather, father, uncle, father-in-law, step father. To keep it simple, we’ll assume two grandfathers, one father, one uncle and one father-in-law. So, five possible targets. 0.18*5=0.9


Passes from lung cancer, emphysema, TB, chest area: This refers directly to the older male. The chances of a male of 65 dying of these diseases is, I believe (my knowledge on medical jargon isn’t too strong, so I could be misreading the report) 1 in 12 for lung problems (not including cancer) and 1 in 3 for heart failure. That makes 0.416 recurring.

BUT this is linked to the mother. So although the medium makes one guess, the sitter invents a second guess. The odds of this inferred guess should, I feel, also be worked out.

Now the guess is for everyone older, male or female, and now refers only to having the lung condition, not dying from it. As far as I can tell, this pushes the probability over one. But it’s possible (if improbable) for the caller to not know anyone older who had chest problems. Let’s call it 0.95

J or G sounding name attached to this: so the person with the J or G sounding name is not specified. It could be anyone. J or G names are so common that, again, the probabilities go over one. So, 0.95 it is.

August or 8th: 1/12, more or less. 0.083 recurring

So, odds of a hit for each guess:

Older male: 0.9
Chest area for one person: 0.416
Chest area for older person: 0.95
J or G: 0.95
August: 0.083

5 guesses, 2 hits

So I just need someone to work out for me, what are the chances of getting ANY two hits out of those options. It doesn’t have to be exactly those hits. After all, the medium isn’t picky about what gets a hit and what doesn’t: just as long as overall the sitter keeps saying “yes”.

[CFLarsen]

Originally posted by Garrette
You are referring to the method of disclosure. My comment was in reference to what is being disclosed.

Correct. However, I don't see why email cannot be used in both cases.

[T'ai Chi]

Originally posted by CFLarsen

But how can you have me on ignore, if you see and respond to my posts??


I had you on my Ignore list, then I'd click to view your individual posts.


You did not have me on ignore.


No, I had you on my ignore list.


Stop playing these juvenile games, T'ai Chi. You can email me at webmaster@skepticreport.com when notifications/updates/etc. occur.


No, I'll PM you (everyone) with updates if/when they occur, as stated long ago.


It's "SkepticReport". Why wouldn't you? It provides a forum where you can publish your research. Where else would you publish it?

Where did I even mention anything about publishing the Excel spreadsheet? Please let me know. :)

[T'ai Chi]

Originally posted by CFLarsen

However, I don't see why email cannot be used in both cases.

If you'd run it differently, you are welcome to start your own study.

[CFLarsen]

Originally posted by T'ai Chi
I had you on my Ignore list, then I'd click to view your individual posts.

You have me on ignore, yet click on my posts?

Originally posted by T'ai Chi
No, I had you on my ignore list.

You have me on ignore, yet click on my posts?

Originally posted by T'ai Chi
No, I'll PM you (everyone) with updates if/when they occur, as stated long ago.

Sorry, you also allowed for email. Please use my email from now on.

Originally posted by T'ai Chi
Where did I even mention anything about publishing the Excel spreadsheet? Please let me know. :)

You are not going to publish the Excel spreadsheet? It will not be made public?

Originally posted by T'ai Chi
If you'd run it differently, you are welcome to start your own study.

Sorry, you also allowed for email.

[T'ai Chi]

Ersby,

Thanks for your constructive comments on things to look for!

Yes, there is certainly a lot of work involved in this. I'll look over these and definitely add most if not all of them to the spreadsheet.


Originally posted by Ersby

Here's the categories I think should be included...


Basic Hit rate (in which a hit is when the sitter responds to something in a positive way. This includes everything, even repeating back information and “Do you understand?”)

No. of misses

No. of unacknowledged statements (a statement made by the medium which is never addressed by the sitter)

No. of “Do you understand”

No. of repeating back information

No. of common sense conclusions (whereby if the medium is told they sold a house, he guesses the house needed a little work and gets a hit. This includes guesses about jewellery and watches for older deceased, and guesses about toys for younger energies)

No. of guesses made repeatedly and adjusted by medium

No. of guesses stretched by sitter to fit something

Underlying hit rate (Basic hit rate minus the DYU count, the repeating count, the common sense count, the adjusted guess count and the stretched hit count)


Names:

No. of name guesses

No. of hits

No. of misses

No. of unacknowledged

No. of full-name guesses referred implicitly to one person (was your father called Michael?)

No. of exact hits for those guesses (ie, the father was called Michael)

No. of stretched hits for those guesses (ie, the son was called Michael)

No. of misses for those guesses

No. of unacknowledged full-name single person guesses

No. of full-name guesses to a range of people (ie, is your grandfather called Michael)

No. of exact hits for those guesses

No. of stretched hits for those guesses

No. of misses for those guesses

No. of unacknowledged full-name range guesses

No. of full-name guesses linked to another guess (ie, I’m getting a problem with the lungs, and there’s a Michael connection too)

No. of exact hits for those guesses (ie, Michael had lung problems)

No. of stretched hits for those guesses (both Michael and lung problems mean something, but are not directly linked)

No. of misses for those guesses (miss on Michael)

No. of unacknowledged full-name guesses linked to another guess

No. of full name guesses with no context (i.e., Who’s Michael?)

No. of exact hits for those guesses (i.e., Michael was a person referred to in the reading before the guess)

No. of stretched hits for those guesses (i.e., Michael is someone not mentioned before)

No. of misses for those guesses

No. of unacknowledged full-name no-context guesses


This would then be followed by the same categories for “Choice of full names” (ie, John, Joe) and “Initials” (Peter, P name).

Then it would further followed up by a measure of the probability of getting a hit particular name, taking into account the range of possible targets.

As you see, the names alone constitute an enormous amount of work if they’re to be done properly. Then of course you must understand that if the medium asks about “Michael” but the sitter replies “Mark” and it is accepted by the medium, then that goes into the “Initial” category. Although the medium guessed only one name, it was understood by both parties that any male M name would be accepted.

And that's only the beginning of what should be examined!

[T'ai Chi]

Claus, I have no problems not updating people who constantly argue, nitpick, and waste my time. I could care less. Take note.

Originally posted by CFLarsen

You have me on ignore, yet click on my posts?
You have me on ignore, yet click on my posts?


Yes, as I've already stated.

It makes your posts stand out more that way actually since all surrounding posts are from posters not on my ignore list. Then I just click to view a single post.


You are not going to publish the Excel spreadsheet? It will not be made public?


I've already mentioned that I will put it up on a webpage when done, not yours. If you're not involved with actually doing anything, why on earth would the final results be put on SkepticalReport.

[TLN]

Originally posted by T'ai Chi
I've already mentioned that I will put it up on a webpage when done, not yours.

When can we expect this?

[T'ai Chi]

Bumping thread.

[CFLarsen]

So, when can we expect this?

[T'ai Chi]

Originally posted by CFLarsen
So, when can we expect this?

In roughly 2 to 3 years?

[T'ai Chi]

To further counter claims of no secrecy in science, nother area in science where there is secrecy is in any experiment where there is blinding.

[CFLarsen]

Originally posted by T'ai Chi
In roughly 2 to 3 years?

How do you calculate that?

Do you mind keeping us informed, e.g. by posting monthly bulletins?

[T'ai Chi]

Originally posted by CFLarsen

How do you calculate that?


Rough approximations.


Do you mind keeping us informed, e.g. by posting monthly bulletins?

I'll update those that are involved with doing the work, and will post to this thread when it is done.

[CFLarsen]

Originally posted by T'ai Chi
Rough approximations.

Based on what, precisely?

Why is it so difficult for you to be precise in your replies? Why do we have to drag everything out of you, wasting a lot of time and effort in doing so?

Originally posted by T'ai Chi
I'll update those that are involved with doing the work, and will post to this thread when it is done.

I have PM'd you that I was interested, and you have acknowledged that. So far, I have heard absolutely nothing from you. I can only assume that you have gotten nowhere in your work.

If I am wrong, could you please live up to your promise and PM me about the progress?

[T'ai Chi]

Originally posted by CFLarsen

Based on what, precisely?


An approximation based on how long I think it will take to complete it. Assuming there are about 2 or 3 hundred transcripts out there and it takes half a week to fully analyze a single transcript. These are probably overestimates.


I have PM'd you that I was interested, and you have acknowledged that. So far, I have heard absolutely nothing from you. I can only assume that you have gotten nowhere in your work.

If I am wrong, could you please live up to your promise and PM me about the progress?

It hasn't even been a month yet and things are still in the planning stage. No transcripts have even been analyzed yet Claus.

I'll PM people when I deem it necessary. If you disagree or would do it differently, you are free to start your own study.

[CFLarsen]

T'ai Chi,

How many transcripts do you have so far?

[T'ai Chi]

Originally posted by CFLarsen
T'ai Chi,

How many transcripts do you have so far?

About 20.

[CFLarsen]

Originally posted by T'ai Chi
About 20.
So, what are you waiting for? Shouldn't you start sending out transcripts to those who signed up??

[T'ai Chi]

Originally posted by CFLarsen

So, what are you waiting for? Shouldn't you start sending out transcripts to those who signed up??

If you disagree or would like to do it differently, you are encouraged to start your own study.

I'm waiting to get more transcripts, especially those by professional cold readers. So far, all the transcripts are those by professional mediums.

[T'ai Chi]

Do people think it is wiser to only analyze the transcripts from live readings?

This will of course create WAY less transcripts to analyze, but more meaningful data could be obtained.

Comments?

[CFLarsen]

T'ai Chi,

How many more transcripts do you need, before something happens?

You seem to be fumbling in the dark.

[Darat]

Bump

[CFLarsen]

It seems that T'ai Chi is more busy opening new threads about his "study", than actually doing it.

[BillHoyt]

Originally posted by CFLarsen
T'ai Chi,

How many more transcripts do you need, before something happens?

You seem to be fumbling in the dark.
What can possibly happen other than the production of a stamp-collecting catalog? Except for T'ai's slip into discussing inferential statistics, he's just collecting data. Because he has not set forth an a priori hypothesis, he cannot then use this data as a mine from which to extract any conclusions. So what remains here is the long, slow, march to the death of the idea by:

o the whole thing shrinking away and shrivelling up,
o announced results simply as a collection of means, counts, variances
o T'ai pulling out a statistically false conclusion.

I think what we're really watching here is T'ai, having finally realized he cannot reach any conclusions via the orignal procedure he described, trying to make this all go away.

[Ed]

Originally posted by BillHoyt



I think what we're really watching here is T'ai, having finally realized he cannot reach any conclusions via the orignal procedure he described, trying to make this all go away.

I, personally, would give it up as a non-starter. The work will be a pain in the ass. Did T'ai mention anything about reliability? It seems to me that you have to have the data coded a few times to assure reliability. As I said, pain in the ass for some descriptive stats.

[CFLarsen]

Originally posted by Ed
Did T'ai mention anything about reliability?

Nobody knows if reliability is dealt with. Secrecy is tight, for one reason only: When one asks T'ai Chi about something, the answer might come (after a tedious series of posts), and then, always fumbling.

"Yeah, sure, it's a concern" which means "Oh, dang, didn't think of that either!"

[T'ai Chi]

Originally posted by BillHoyt

Except for T'ai's slip into discussing inferential statistics,


I haven't done that, for reasons already discussed. I said 'for the data we examined'. You didn't on don't understand, that isn't my problem.

Bill, I'm curious, do you still believe that it is only called skew? :D :D :D
(avoid this question as usual!)


o the whole thing shrinking away and shrivelling up,
o announced results simply as a collection of means, counts, variances
o T'ai pulling out a statistically false conclusion.


Neither.

A resource, in the form of a spreadsheet, of meaningful measures from the unedited live transcripts from professional cold readers and professional mediums, available for anyone in the skeptical and believer community to study.

Again, you are free to offer your ideas on how to examine these. I predict that you'll have tons of comments, but only AFTER the study is done. That is the only thing you appear to be able to offer.

And no Bill, it will not go away no matter what you, Claud, or anyone else say to misrepresent the study or my supposed intentions.

So, is skew only called skew? Yes or No? Will Bill admit he was wrong? Nope.

[CFLarsen]

Originally posted by T'ai Chi
A resource, in the form of a spreadsheet, of meaningful measures from the unedited live transcripts from professional cold readers and professional mediums, available for anyone in the skeptical and believer community to study.

You have 25 unedited live transcripts? From where?

[T'ai Chi]

Originally posted by CFLarsen

You have 25 unedited live transcripts? From where?

How did you get "25" from "about 20"?

Must be that 'new math'.

Anyway, from various live talkshows.

[CFLarsen]

Originally posted by T'ai Chi
How did you get "25" from "about 20"?

Must be that 'new math'.

Anyway, from various live talkshows.

Whatever. Are you going to send them to me?

[T'ai Chi]

Originally posted by CFLarsen


Whatever. Are you going to send them to me?

WTF would I do that? You turn about 20 into 25, who knows how you'd transform more data.

[True-Gossiper]

errr.. this discussion is going nowhere..

[CFLarsen]

Originally posted by T'ai Chi
WTF would I do that? You turn about 20 into 25, who knows how you'd transform more data.

I thought you said 25, I was wrong.

Can I have them for further reference? You know that I keep threads, data, etc., so why not these transcripts?

I could also make them available for others. I'd be happy to provide server space for them.

[T'ai Chi]

Originally posted by CFLarsen


I thought you said 25, I was wrong.


Its a miiiirrrraaaacccclllleeeeee!


Can I have them for further reference? You know that I keep threads, data, etc., so why not these transcripts?

I could also make them available for others. I'd be happy to provide server space for them.

CNN's Larry King Live site has some of these. I'm going to set up a webpage in the next month or so. I'll put them up, along with some corrections and critiques of SkepticReport.com's articles that I've been working on, as well as some other things.

[CFLarsen]

Originally posted by T'ai Chi
Its a miiiirrrraaaacccclllleeeeee!

Not at all. If I am wrong, I admit it.

Originally posted by T'ai Chi
CNN's Larry King Live site has some of these. I'm going to set up a webpage in the next month or so. I'll put them up, along with some corrections and critiques of SkepticReport.com's articles that I've been working on, as well as some other things.

I don't see why you cannot just email them to me. But if you want to put them on your own site, go ahead. Shall we say I'll check on you June 1st?

[Ed]

Originally posted by T'ai Chi


Its a miiiirrrraaaacccclllleeeeee!



CNN's Larry King Live site has some of these. I'm going to set up a webpage in the next month or so. I'll put them up, along with some corrections and critiques of SkepticReport.com's articles that I've been working on, as well as some other things. [/B]

Wait a moment. Is this going to be like the "analysis" that you backed out of doing of that silly paper?

[T'ai Chi]

Originally posted by Ed


Wait a moment. Is this going to be like the "analysis" that you backed out of doing of that silly paper?

I posted my comments on it.

If your brain thinks analysis equates with paper, that is your problem.

My favorite from that thread was Jr., who after avoiding pertinent statistical issues tried to go the copyright violation route. After getting permission that it was ok to post the person's comments, Jr. still hasn't responded. Figures.

[T'ai Chi]

Originally posted by CFLarsen


Not at all. If I am wrong, I admit it.


That's a hoot.


I don't see why you cannot just email them to me. But if you want to put them on your own site, go ahead. Shall we say I'll check on you June 1st?

Since I don't know exactly when I'll get a webpage all set up, I'll let people know instead of setting a date.

Will you post a correction to your % agreement article? Will you post a followup on the 65% CI issue?

Let us know.

[CFLarsen]

Originally posted by T'ai Chi
That's a hoot.

No, that's evidence. Something for you to ignore, though it won't go away.

Originally posted by T'ai Chi
Since I don't know exactly when I'll get a webpage all set up, I'll let people know instead of setting a date.

Of course you won't! I will ask you once a week to see how far you have gotten. Just so you don't think that you will get away with letting it slide into oblivion.

Originally posted by T'ai Chi
Will you post a correction to your % agreement article? Will you post a followup on the 65% CI issue?

Let us know.

Email your complaints to the authors, like everyone else. You do not get preferential treatment.

[T'ai Chi]

Originally posted by CFLarsen

Email your complaints to the authors, like everyone else. You do not get preferential treatment.

Um, you are the author of the % agreement article, dood.

:rolleyes:

[CFLarsen]

Originally posted by T'ai Chi
Um, you are the author of the % agreement article, dood.

:rolleyes:

Email.

[T'ai Chi]

Originally posted by CFLarsen

Email.

I have to email to make my thoughts known?

Absurd. You.

[CFLarsen]

Originally posted by T'ai Chi
I have to email to make my thoughts known?

Absurd. You.

Not at all. Nobody gets preferential treatment. Not even you.

[T'ai Chi]

Originally posted by CFLarsen


Not at all. Nobody gets preferential treatment. Not even you.

You were the author of that article. You already know my thoughts on it and have read my improved analysis (where an improved calculation of the % agreement was shown) that I posted in this forum. Now you are saying that I have to email it before you've really read it.

Play your games elsewhere.