Cold Reading Demos at TAM2

[BillHoyt]

Read back over my posts in this thread. I stated on page 16:
You have now pointed me to that general post. Whoopee. Care to connect it to JE? That is all I am asking. I asked in the other thread too, but you never did answer me. Are you going to explain it or not?

Are you going to answer the other points and questions I have asked in this thread?

You deign to tell me the method is flawed and the model is flawed, but you never understood the explanation? The connection you ask for is obvious and embodied in the explanation:

The definition of a Poisson distribution is based on that of a Poisson process. A Poisson process is one that satisfies the following:


1. The changes (or events) that result from the process can be grouped into nonoverlapping intervals.

True for JEs name guess utterances. They can be grouped into N nonoverlapping intervals, where N is the number of name guess utterances made by JE.

2. The numbers of changes (or events) in the nonoverlapping intervals are independent from one another.
True again. JE's choice to say a "J" or "non-J" are independent.

3. This independence holds for all intervals.
Again, true. The independence here is similar to that of Poisson models of radioactive decay.

4. The probability of exactly one change (or event) in a sufficiently small interval, h = 1/n equals n*p , where p is the probability of one change (or event) and n is the number of trials.
Again, true. The p here for "J" comes from the census data. The n comes from the total number of guesses made.

5. The probability of more than one change (or event) in a sufficiently small interval, h, is essentially 0.
As the intervals chosen are the guesses, this is obviously true. Here, in fact, the probability is almost exactly 0.

The Poisson distribution results when such a process occurs over n trials

 

[Thanz]

Mr. Hoyt -

I have a problem with points 2, 3 and 4.

When JE says "A J connection, like....." the following events are not independent from the first - J connection. It seems to me that the examples following J connection do not have the same independence as the three separate readings. I think this was doge69's problem as well.

The first event, J, has a probability of p from the census data. The second event, however, does not have the same probability. It is far more likely to also be a J than any other letter.

 

[BillHoyt]

Mr. Hoyt -

I have a problem with points 2, 3 and 4.

When JE says "A J connection, like....." the following events are not independent from the first - J connection. It seems to me that the examples following J connection do not have the same independence as the three separate readings. I think this was doge69's problem as well.

The first event, J, has a probability of p from the census data. The second event, however, does not have the same probability. It is far more likely to also be a J than any other letter.

And this differs from radioactive decay how?

 

[Thanz]

Well, you got me there Hoyt. I have no idea how it differs from radioactive decay, as I know nothing about radioactive decay.

 

[T'ai Chi]

Radar and leopards are fine, but coins are right out? Of course....

Coins are not leopards and radars Thanz!! Duh!!!!

It may move to the mean if you increased the sample size by including more transcripts. All you have done is count multiples of the same guess

Thanz, you're right on here. It seems to me to be the problem of what is called 'pseudoreplication', and was a major problem in ecology in the 70's and 80's, where researchers were studying, say, 2 lakes, and to try and make up for their n = 2, they would measure a lot of things in the lakes. Well, they can measure a million things in the lake, but they still only have n = 2.

Our experimental unit with mediums seems to be the transcript. Taking as many counts as we can from the transcript shouldn't increase our understanding, rather only having more transcripts will.

 

[T'ai Chi]

In other circumstances, where we get the whole bird in pieces, we just count the one bird. "J connection, like John or Joe" is like finding a dead bird with the wings detached, but right beside the bird. We don't count 3 birds. We count one.

Exactly Thanz, which is another way of saying Bill's assumption of independence (which is, warning, technical term: hella important) is not met.

 

[T'ai Chi]

Thanz,
What, no statistics?

 

[T'ai Chi]

Not at all. This is about statistics, Thanz. Please use statistics in a discussion about statistics.

Funny, I don't think you required that in the DAT thread.

By using that, you have agreed that Hoyt's data are sound. Why else use it??

He could use it to check work, or to demonstrate a point for example.

 

[BillHoyt]

Our experimental unit with mediums seems to be the transcript. Taking as many counts as we can from the transcript shouldn't increase our understanding, rather only having more transcripts will.

Amazing, Tr'oll. We have one datum per transcript. Hello?

How do the JE utterances differ from radioactive decay, sir?

 

[T'ai Chi]

Originally posted by CFLarsen (edited by T'ai Chi)


◊◊◊◊◊◊◊◊! By arguing from the data, using the calculator, you have endorsed his data.

Why use the same calculator, if you thought the data was bad? Were all your posts about the Poisson calculator simply a waste of time?

◊◊◊◊◊◊◊◊, Thanz.

Saved for posterity.

 

[BillHoyt]

Exactly Thanz, which is another way of saying Bill's assumption of independence (which is, warning, technical term: hella important) is not met. [/B]

And this differs from radioactive decay how?

 

[Thanz]

And this differs from radioactive decay how?

Um, how are JE guesses the same as radioactive decay?

 

[BillHoyt]

Um, how are JE guesses the same as radioactive decay?

One collision can result in two or three or more. There are markov chains within the mass under investigation. Yet, the Poisson model is perfectly applicable until the mean gets too high, and one must switch to normal.

 

[T'ai Chi]

Amazing, Tr'oll. We have one datum per transcript. Hello?

You have many datums, but the transcripts are independent (experimental units, even though we aren't doing an experiment), not the readings in the transcripts, for reasons Thanz and others have already pointed out.

For example, if JE sees/senses/etc. a grandfather, says he's getting 'a J name, John, Joe, Joseph, Jimmy' there is no independence of these guesses as you'd like to believe. Because JE knows/believes it is a J name, the last 3 guesses are all dependent on the first and on each other.

How do the JE utterances differ from radioactive decay, sir?

Radioactive decay is the process by which some atoms spontaneously disintegrate, emitting both particles and energy as they transform into different, more stable atoms. JE is a medium claiming to be able to get messages from the other side. We can read these 'messages' in transcripts.

 

[BillHoyt]

Radioactive decay is the process by which some atoms spontaneously disintegrate, emitting both particles and energy as they transform into different, more stable atoms. JE is a medium claiming to be able to get messages from the other side. We can read these 'messages' in transcripts.

Can't answer the question statistically, Tr'oll? I'm getting an M connection. Markov? Moron? Muttonhead?

 

[CFLarsen]

Thanz,

If your comparison was based on flawed data (in your opinion), what could such a comparison possibly prove?

If you don't want to answer, just say so.

 

[Thanz]

One collision can result in two or three or more. There are markov chains within the mass under investigation. Yet, the Poisson model is perfectly applicable until the mean gets too high, and one must switch to normal.

It would be good if you could be a little bit more specific here. What is a markov chain? Does it make any difference in the decay if it is a primary collision or a secondary collision? What exactly are you studying with the Poisson distribution?

Sorry, but as I said, I don't really know anything about radioactive decay.

 

[T'ai Chi]

And this differs from radioactive decay how?

Individual nuclear decays are purely random events. JE's first guess of a name might be, but his followups aren't, as one already knows they will also be names with the same letter.

 

[T'ai Chi]

I'm getting an M connection. Markov? Moron? Muttonhead?

Mm, that made me hungry for lunch!

 

[Thanz]

Thanz,

If your comparison was based on flawed data (in your opinion), what could such a comparison possibly prove?

If you don't want to answer, just say so.

I'm sorry, but I just don't know what you are driving at. What is the flawed data that you are talking about here?