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