Here's the story...

http://cosmiclog.msnbc.msn.com/_news/201......-psychic

Right now, the only thing Daryl Bem has is a "paper." An interesting paper, but a paper nonetheless… just like thousands of students produce on a daily basis. Though I would like to believe that scientists might someday acknowledge the existence of the supernatural, their egos probably won't allow that to happen.

The chief problem I see with the precognitive section of his testing is the sample size. With a sample size of 100 to 150 subjects, I don't think he can legitimately make the claims regarding chance that he does. Given that his percentage of correct identification dropped from 53.1 to 51.7 just by adding just 50 subjects, I contend that the percentage will continue to drop as the sample size increases and will likely come to rest on 50/50 around 800 to 1000 subjects.

It was still a great piece to read, especially since it caught the attention of other scientists, even if they couldn't duplicate the results. I'd like to hear more about how closely Bem followed the scientific method in conducting his experiments.

I'm a social psychologist, Michael, and quite offended by your comments about our field, the fact that we just "write papers" and your lack of understanding regarding statistics in general. First of all, Daryl Bem is a pre-eminent social psychologist most famous for developing self-perception theory, and ranks just a tier below Festinger, Zimbardo, and Milgram in our field in impact.

"The chief problem I see with the precognitive section of his testing is the sample size. With a sample size of 100 to 150 subjects, I don't think he can legitimately make the claims regarding chance that he does. Given that his percentage of correct identification dropped from 53.1 to 51.7 just by adding just 50 subjects, I contend that the percentage will continue to drop as the sample size increases and will likely come to rest on 50/50 around 800 to 1000 subjects."

You do realize that statistical results are meant not to describe the sample tested but to generalize the findings from that sample to the population at large, do you not? Your speculation that the findings would be less significant over a much larger sample size is unfounded. In fact, if he only was able to obtain the significance results he achieved with a sample size in the range you suggested, it would mean that the difference between conditions is relatively negligible. A paper with sample sizes in that range would be viewed as too negligible to publish in a quality peer-reviewed journal like JPSP. The effect sizes he reported in several of his findings are quite large in statistical terms, not easily dismissed as occurring out of luck (i.e., they are not small differences in statistical terms). That he was able to demonstrate his effect with a sample size of 100 or so suggests that his effect is quite power.

You also do realize that in psychological research of any merit, you can't just keep replicating studies in the same exact manner. Generally, you alter your methodology from one study to the next to rule out alternative explanations. You can't compare percentages across studies using different methodologies as any kind of evidence for a decrease in the power of an effect over a larger sample size, like you attempted to do. Additionally, the percentages reported are pretty much the least important piece of statistical information anyone with a background in statistics would extract from an article--what's more important is that each of the differences reported has less than a 5% likelihood of being different in the way determined due to chance alone, and the effect sizes reported detail what percentage of the variance in the outcome variable is due to the independent variable alone, and what percentage is due to other factors not accounted for.

Finally, it is not "just a paper... like thousands of students produce on a daily basis." Here's a little perspective for you. Roughly 15% of people applying to doctoral programs in social psychology in the United States are accepted to a program any year. There are roughly 100 to 120 schools offering a Ph.D in this field in the US, with probably means there are about 150 students being accepted into a Ph.D. program in a given year. If accepted, you then work between 90 and 100 hours a week for 5-7 years to complete your Ph.D. However, that alone doesn't indicate that you'll get a good position. Here there is roughly a 10% chance of someone who applies for a position as a professor at a research 1 school, like Cornell, to get any position of that nature. On average, there are maybe 15 to 20 openings in the country in a given year in such a position. With regards to what is present in this "just a paper like thousands of students write everyday": You have to carefully design each study based on a considerable literature review that takes several months if you want to have anywhere near the level of knowledge in your topic of interest necessary to be published in JPSP, you have to carefully plan your research methods based on your years of training, you have to go through the institutional review board process, which involves considerable paperwork and time, you have to have the resources in graduate students and undergraduates necessary to run the study. Then, you have to hope your study generates compelling enough results to support your hypothesis which occurs far less often than you'd think because of (surprise, surprise) the relative objectivity of strenuous statistical testing. I personally have about a 20 to 30% success rate with any study I design turning out as I'd hoped, even with careful planning drawn from the findings of previous studies. Then, you have to string together several more studies that also support your predictions if you want a decent publication. The time that it takes to start conducting research on a topic to where you've strung together enough supportive findings to get the paper published in a good journal often takes several years. Finally, you write up your paper with your 3 or so supportive studies and submit it for publication. The Journal of Personality and Social Psychology is the top journal in the field of social psychology. Only about 10% of the articles submitted to JPSP are accepted for publication there. Your ignorance of the amount of time it takes to plan and conduct experiments for a paper of this nature, coupled with the minimal probability of even being in the position that Daryl Bem is in is truly insulting, beyond your complete lack of understanding in statistics.

I believe that if science doesn't prove the existence of psychic ability now, doesn't mean it won't in the future. Sometimes all science does is confirm what people already know.

Have to agree with cpbartak on this. Dismissing Bem's work as "just another paper - like thousands of students produce on a daily basis" was inaccurate and ill informed.

Just as a side note - Dr. Bem is also an accomplished mentalist and an early member of the P.E.A.

Good thoughts,

Bob

"their egos probably won't allow that to happen."

What a crock. Science is about searching for reality--what really is, as opposed to what you may wish it to be. Nothing would please the scientific community more than establishing psychic abilities as genuine. Science is afraid of nothing, especially change. It would be wonderful to hear the cries..."at LAST we have evidence!"

Benny

Cpbartak -

Thanks for the detailed explanation! You obviously know your "stuff".

Many years ago Richard Dawkins came up with an interesting theoretical argument against the scientific validity of psychic ability.
If this ability were genuine, it would give significant advantage to the individual, and these psychic traits would spread throughout the population.
This has clearly not happened.

Assuming psychic traits are genetic... Are they?

I knew this would happen.

in deet it is allways a problem when the ego / wishfull thinking subsitutes facing facts/plausibility or at least a reason based theroie..

In regard to the thread title: If we are psychic, and that is specifically defined by those who claim it, then I see no reason why science cannot prove it (to the extent that science proves anything). Whatever the definition of psychic ends up being, I assume those who claim to have the ability make that claim based on something demonstrable. If it is demonstrable, then it is testable/replicable. If it is not demonstrable then there is no reason to assume one has the ability.

Regarding Bem's paper in the OP, I became aware of this attempt at replication from a JREF thread:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1699970

The JREF thread has some excellent discussion, pro and con, of the paper and should be of interest to some here. This is the link:http://forums.randi.org/showthread.php?t=188366

The discussion is primarily among those qualified to look at the merits of the paper as opposed to opinionated amateurs like me, so I recommend it.

ETA: Bob, I didn't know Bem is a mentalist. Thanks for that tidbit.

I knew Bev knew this would happen.

this is the proof for psychic abilitys do exist.. ;-)

The ability to read minds has caused me to lose any desire to reproduce.

Bob

Lol- and the ability of women I've met in bars to read minds has not allowed me to...well, so many things.

"Assuming psychic traits are genetic... Are they?"

One first has to establish the existence of "psychic traits" before you can establish the latter. First things first.

Benny

And that is what the guy is doing or has done.

As for the genetic deal only a fool would not know the answer to that.

But this place draws them like rats to a full trash bin.

Jim

H.O.A-X

It is more than half a century now since scientists began trying to find psychic abilities. In fact JB Rhine began his research in 1931. that's 79 years of searching without finding anything. This new paper is more of the same. No matter how eminent the author, the results will not be replicated. The reason is simple; you can't find what is not there.

Your post did make me smile.

It should not but it did.

Thanks,

Jim

H.o.A-X