Print Page - Interesting Research by Cornell Psychology Prof

Title: Interesting Research by Cornell Psychology Prof
Post by: Ken70 on January 17, 2011, 12:31:26 PM

Research that shows in some cases humans can anticipate and act on future events before they occur...fascinating.

http://dbem.ws/FeelingFuture.pdf

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Robb on January 17, 2011, 05:57:42 PM

Quote from: Ken70Research that shows in some cases humans can anticipate and act on future events before they occur...fascinating.

http://dbem.ws/FeelingFuture.pdf

Saw this a while ago and immediately thought, "I bet half the folks on eLynah could debunk his statistics..."

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Beeeej on January 17, 2011, 06:06:22 PM

Quote from: Robb
Quote from: Ken70Research that shows in some cases humans can anticipate and act on future events before they occur...fascinating.

http://dbem.ws/FeelingFuture.pdf
Saw this a while ago and immediately thought, "I bet half the folks on eLynah could debunk his statistics..."

I knew that's what you were going to say.

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Ken70 on January 17, 2011, 09:08:34 PM

Quote from: Robb
Quote from: Ken70Research that shows in some cases humans can anticipate and act on future events before they occur...fascinating.

http://dbem.ws/FeelingFuture.pdf
Saw this a while ago and immediately thought, "I bet half the folks on eLynah could debunk his statistics..."

I bet they can't, nor could you. Do a little research and you'll see the peer review found nothing to take exception to. Immediate thoughts aren't as reliable as considered ones.

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Jacob '06 on January 17, 2011, 09:42:17 PM

Quote from: Ken70
Quote from: Robb
Quote from: Ken70Research that shows in some cases humans can anticipate and act on future events before they occur...fascinating.

http://dbem.ws/FeelingFuture.pdf
Saw this a while ago and immediately thought, "I bet half the folks on eLynah could debunk his statistics..."

I bet they can't, nor could you. Do a little research and you'll see the peer review found nothing to take exception to. Immediate thoughts aren't as reliable as considered ones.

It got through peer review, but that is not the be all end all of it being correct. In fact, most of the people who peer reviewed it likely had little expertise in statistics since it is a psychology journal and would have gone to psychologists. In fact there is debate about the validity of the statistics used in the article: http://www.nytimes.com/2011/01/11/science/11esp.html

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Robb on January 17, 2011, 10:36:48 PM

Quote from: Ken70
Quote from: Robb
Quote from: Ken70Research that shows in some cases humans can anticipate and act on future events before they occur...fascinating.

http://dbem.ws/FeelingFuture.pdf
Saw this a while ago and immediately thought, "I bet half the folks on eLynah could debunk his statistics..."

I bet they can't, nor could you. Do a little research and you'll see the peer review found nothing to take exception to. Immediate thoughts aren't as reliable as considered ones.

I've read several articles about it, thankyouverymuch (including the one Jacob posted). They found 53% "heads" in one of the MANY studies that they conducted - the others all came up "negative" (i.e. only 50% correct guesses). I'm guessing a meta-analysis across all the studies would suggest that it would actually be fairly improbable that ALL the studies would come out to exactly 50%. In fact, he found the effect to be "significant" with a p-value of 0.01. In other words, there's a 1% probability that what he encountered would have happened by chance - big deal.

He was trying to test so many different things with a limited number of students and trials that there weren't all that many actual trials for the "erotic" pictures that showed the 53% result. All the news articles of course report that "1000 students were used" in the study, but that is highly misleading. This is why his significance is merely 0.01 - there weren't enough trials with the erotic pictures to be any more confident than that. It's the probability of flipping 7 heads in a row - unlikely? Sure, but in reality it should happen almost 1% of the times that you try it.

And then he has to go and throw statements around like "What I showed was that unselected subjects could sense the erotic photos, but my guess is that if you use more talented people, who are better at this, they could find any of the photos." Uh-huh. And a p-0.01 significance in one small portion of the overall results supports this conclusion how?

Show me a study with a p-value in the 1e-6 range and I might start to have second thoughts...

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: ugarte on January 18, 2011, 12:33:27 AM

I predict that this study will be shown to be complete bullshit, thereby proving the study correct after all.

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: jtwcornell91 on January 18, 2011, 07:39:08 AM

Quote from: Jacob '06
Quote from: Ken70
Quote from: Robb
Quote from: Ken70Research that shows in some cases humans can anticipate and act on future events before they occur...fascinating.

http://dbem.ws/FeelingFuture.pdf
Saw this a while ago and immediately thought, "I bet half the folks on eLynah could debunk his statistics..."

I bet they can't, nor could you. Do a little research and you'll see the peer review found nothing to take exception to. Immediate thoughts aren't as reliable as considered ones.

It got through peer review, but that is not the be all end all of it being correct. In fact, most of the people who peer reviewed it likely had little expertise in statistics since it is a psychology journal and would have gone to psychologists. In fact there is debate about the validity of the statistics used in the article: http://www.nytimes.com/2011/01/11/science/11esp.html

The rebuttal article (http://dl.dropbox.com/u/1018886/Bem6.pdf) is a rather nice read, and delves into the larger issues related to statistical analysis as it's used in psychology.

ETA: this is indeed interesting (although I think not for the reasons Ken meant), and I hadn't seen it; thanks for bringing it to my attention.

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Jim Hyla on January 19, 2011, 08:04:56 AM

Quote from: Robb
Quote from: Ken70
Quote from: Robb
Quote from: Ken70Research that shows in some cases humans can anticipate and act on future events before they occur...fascinating.

http://dbem.ws/FeelingFuture.pdf
Saw this a while ago and immediately thought, "I bet half the folks on eLynah could debunk his statistics..."

I bet they can't, nor could you. Do a little research and you'll see the peer review found nothing to take exception to. Immediate thoughts aren't as reliable as considered ones.
I've read several articles about it, thankyouverymuch (including the one Jacob posted). They found 53% "heads" in one of the MANY studies that they conducted - the others all came up "negative" (i.e. only 50% correct guesses). I'm guessing a meta-analysis across all the studies would suggest that it would actually be fairly improbable that ALL the studies would come out to exactly 50%. In fact, he found the effect to be "significant" with a p-value of 0.01. In other words, there's a 1% probability that what he encountered would have happened by chance - big deal.

He was trying to test so many different things with a limited number of students and trials that there weren't all that many actual trials for the "erotic" pictures that showed the 53% result. All the news articles of course report that "1000 students were used" in the study, but that is highly misleading. This is why his significance is merely 0.01 - there weren't enough trials with the erotic pictures to be any more confident than that. It's the probability of flipping 7 heads in a row - unlikely? Sure, but in reality it should happen almost 1% of the times that you try it.

And then he has to go and throw statements around like "What I showed was that unselected subjects could sense the erotic photos, but my guess is that if you use more talented people, who are better at this, they could find any of the photos." Uh-huh. And a p-0.01 significance in one small portion of the overall results supports this conclusion how?

Show me a study with a p-value in the 1e-6 range and I might start to have second thoughts...

Show me a biologic study with those stats and I'll... ,well I don't know what I'd do, other than say I've never seen anything like it. Really, I haven't yet read the rebuttal, but a p value of .01 in medicine is good. It would take extraordinary numbers of subjects to really crank the p down, I think.:-)

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Robb on January 19, 2011, 10:09:46 AM

Quote from: Jim Hyla
Quote from: Robb
Quote from: Ken70
Quote from: Robb
Quote from: Ken70Research that shows in some cases humans can anticipate and act on future events before they occur...fascinating.

http://dbem.ws/FeelingFuture.pdf
Saw this a while ago and immediately thought, "I bet half the folks on eLynah could debunk his statistics..."

I bet they can't, nor could you. Do a little research and you'll see the peer review found nothing to take exception to. Immediate thoughts aren't as reliable as considered ones.
I've read several articles about it, thankyouverymuch (including the one Jacob posted). They found 53% "heads" in one of the MANY studies that they conducted - the others all came up "negative" (i.e. only 50% correct guesses). I'm guessing a meta-analysis across all the studies would suggest that it would actually be fairly improbable that ALL the studies would come out to exactly 50%. In fact, he found the effect to be "significant" with a p-value of 0.01. In other words, there's a 1% probability that what he encountered would have happened by chance - big deal.

He was trying to test so many different things with a limited number of students and trials that there weren't all that many actual trials for the "erotic" pictures that showed the 53% result. All the news articles of course report that "1000 students were used" in the study, but that is highly misleading. This is why his significance is merely 0.01 - there weren't enough trials with the erotic pictures to be any more confident than that. It's the probability of flipping 7 heads in a row - unlikely? Sure, but in reality it should happen almost 1% of the times that you try it.

And then he has to go and throw statements around like "What I showed was that unselected subjects could sense the erotic photos, but my guess is that if you use more talented people, who are better at this, they could find any of the photos." Uh-huh. And a p-0.01 significance in one small portion of the overall results supports this conclusion how?

Show me a study with a p-value in the 1e-6 range and I might start to have second thoughts...
Show me a biologic study with those stats and I'll... ,well I don't know what I'd do, other than say I've never seen anything like it. Really, I haven't yet read the rebuttal, but a p value of .01 in medicine is good. It would take extraordinary numbers of subjects to really crank the p down, I think.:-)

That's exactly the point - for such an extraordinary claim, we should *expect* extraordinary results to confirm it.

JTW - great rebuttal paper. And just for Beeeej, it cited Jeffreys (and some other guy named Jefferys). Can it get any better?

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Trotsky on January 19, 2011, 11:14:07 AM

My favorite blog has a wonderful discussion (http://www.johnreilly.info/08Jan11.htm#psi) of this.

QuoteNonetheless, in reading the rebuttal paper, I got the impression I often get from skepticist literature: the psi question struck some hidden nerve, and the argument is not really about what it purports to be about. It seems that "Feeling the Future" is not just wrong. It is methodologically, metaphysically, and statistically wrong. Indeed, its wrongness is so overdetermined that it is a wonder that the manuscript did not fall like neutron-star material to the center of the Earth

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Jeff Hopkins '82 on January 19, 2011, 12:34:58 PM

Quote from: TrotskyMy favorite blog has a wonderful discussion (http://www.johnreilly.info/08Jan11.htm#psi) of this.

QuoteNonetheless, in reading the rebuttal paper, I got the impression I often get from skepticist literature: the psi question struck some hidden nerve, and the argument is not really about what it purports to be about. It seems that "Feeling the Future" is not just wrong. It is methodologically, metaphysically, and statistically wrong. Indeed, its wrongness is so overdetermined that it is a wonder that the manuscript did not fall like neutron-star material to the center of the Earth

Wouldn't the center of the earth fall into the center of the neutron star material?

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Roy 82 on January 30, 2011, 08:25:25 PM

Bem on Colbert last Thursday.

http://www.colbertnation.com/the-colbert-report-videos/372474/january-27-2011/time-traveling-porn---daryl-bem

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Jeff Hopkins '82 on January 31, 2011, 08:24:26 AM

Quote from: Roy 82Bem on Colbert last Thursday.

http://www.colbertnation.com/the-colbert-report-videos/372474/january-27-2011/time-traveling-porn---daryl-bem

Saw that. Didn't exactly help his "cause", did it?

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Robb on January 31, 2011, 09:49:30 AM

Quote from: Jeff Hopkins '82
Quote from: Roy 82Bem on Colbert last Thursday.

http://www.colbertnation.com/the-colbert-report-videos/372474/january-27-2011/time-traveling-porn---daryl-bem

Saw that. Didn't exactly help his "cause", did it?

Not for me. 53% is "significant" because that's the amount of the popular vote Obama got? Funny, I don't remember studying the "one-tailed Obama test" in any of my stats classes...

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: KeithK on January 31, 2011, 06:15:36 PM

Quote from: Robb
Quote from: Jeff Hopkins '82
Quote from: Roy 82Bem on Colbert last Thursday.

http://www.colbertnation.com/the-colbert-report-videos/372474/january-27-2011/time-traveling-porn---daryl-bem

Saw that. Didn't exactly help his "cause", did it?
Not for me. 53% is "significant" because that's the amount of the popular vote Obama got? Funny, I don't remember studying the "one-tailed Obama test" in any of my stats classes...

53% would be pretty damn significant if he had as many samples as there were votes in 2008.

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Jeff Hopkins '82 on February 01, 2011, 09:04:34 AM

Quote from: KeithK
Quote from: Robb
Quote from: Jeff Hopkins '82
Quote from: Roy 82Bem on Colbert last Thursday.

http://www.colbertnation.com/the-colbert-report-videos/372474/january-27-2011/time-traveling-porn---daryl-bem

Saw that. Didn't exactly help his "cause", did it?
Not for me. 53% is "significant" because that's the amount of the popular vote Obama got? Funny, I don't remember studying the "one-tailed Obama test" in any of my stats classes...
53% would be pretty damn significant if he had as many samples as there were votes in 2008.

And the whole discussion came down to: Weak statistics show that people have ESP because they anticipated porn appearing. Did it have to be porn? Besides the fact that the statistics don't seem overwhelming, the fact that it was porn really makes the study look somewhat silly, or at least childish.

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Robb on February 01, 2011, 09:36:04 AM

Quote from: Jeff Hopkins '82
Quote from: KeithK
Quote from: Robb
Quote from: Jeff Hopkins '82
Quote from: Roy 82Bem on Colbert last Thursday.

http://www.colbertnation.com/the-colbert-report-videos/372474/january-27-2011/time-traveling-porn---daryl-bem

Saw that. Didn't exactly help his "cause", did it?
Not for me. 53% is "significant" because that's the amount of the popular vote Obama got? Funny, I don't remember studying the "one-tailed Obama test" in any of my stats classes...
53% would be pretty damn significant if he had as many samples as there were votes in 2008.

And the whole discussion came down to: Weak statistics show that people have ESP because they anticipated porn appearing. Did it have to be porn? Besides the fact that the statistics don't seem overwhelming, the fact that it was porn really makes the study look somewhat silly, or at least childish.

Apparently, it did - the other 8 studies (that didn't use porn) didn't find anything...

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: Trotsky on February 01, 2011, 11:05:38 AM

Quote from: RobbApparently, it did - the other 8 studies (that didn't use porn) didn't find anything...

Occam's Razor. We are always anticipating porn.

Title: Re: Interesting Research by Cornell Psychology Prof
Post by: ugarte on February 01, 2011, 12:35:33 PM

Quote from: Trotsky
Quote from: RobbApparently, it did - the other 8 studies (that didn't use porn) didn't find anything...
Occam's Razor. We are always anticipating porn.

This. Or we are just optimistic by nature.

ELynah Forum

General Category => John Spencer Is Dead => Topic started by: Ken70 on January 17, 2011, 12:31:26 PM