An experiment was set up where people were shown a video of someone guessing cards and doing well at it. Afterwards they were split into 4 groups of 40 and asked questions including one about the validity of ESP. The four groups were split like this.

Condition 1: Informed that 25 percent of the public believes in ESP, but the scientific community rejects the concept.

Condition 2: Told that more than 90 percent of the public believes in ESP, but the scientific community considers it bogus.

Condition 3: Told that 25 percent of the public believes in ESP, and the scientific community is becoming more open to the idea.

Condition 4: Informed that more than 90 percent of the public believes in ESP, and the scientific community is beginning to warm to the possibility it is real.


The interesting part of their results were that people were MORE likely to except ESP in Condition 1 than in Condition 3. This clearly flies in the face of conventional wisdom. So two questions

1. Do you think the U.S. population as a whole is becoming distrustful of scientists? If so, what kind of implications do you think it would have?

2. Do you have any problems with the design of experiment?

Here is a link to the article: http://www.alternet.org/module/printversion/146552

Distrustful of the "scientific community" (the phrase apparently used by the examiners)? Probably.

Distrustful of the methods of science itself? No.

I don't think people even know what that is.

Here is a link to the paper: http://www.uiowa.edu/~grpproc/crisp/crisp15_3.pdf

I think that science is less "trusted" when it seems to be chasing "random" types of studies. For example, if I heard there was a study about a link between soy sauce and intelligence, I would scoff at it even before knowing anything more about it, and be biased towards doubting the validity of the study.

In the same way, I sort of think that if you hear that people reject ESP, but you're told that science is starting to accept it, you might be inclined to discount the information the scientists have (root against their 'success') because it just seems odd that science would affirm such a thing.

It's also possible that the people in the study (college students) might have "seen through" this particular study and the result reflects a different type of bias -- sort of like a gullibility test. If you're told that scientists believe in ESP but people in general do not, you might shade your confidence down a notch just so that you don't feel like you're being led around by the latest "scientific fads."

I'm not sure I would call that a "design flaw" but rather just a confounding variable.

2. As long as the scale was big enough to weed out variance seems fine if done properly.

1. Yeah and none they have been since after ww2 amirite?

I think I found one source of bias in the study:
If the highlighted means they tossed the answers of people who didn't believe what they were told about the "scientific community" becoming more open to the possibility of ESP, then I suspect they'd be more likely to be throwing out answers of people who were likely to give very low scores (i.e., people who don't believe in ESP after seeing the video). It's also possible that people who believed the video was faked would be more likely to give low scores, but they could be in any of the four groups so that's not really a source of bias.)

An underpinning of my suspicion that this is a source of bias is my guess that people who elt so strongly that they were being lied to that they had to be tossed were much more likely to have a problem with the obviously silly claim that science is beginning to accept the possibility that ESP is real than the correct claim that its not or either claim about popular belief, and that people who realize this is bunk are likely to be nonbelievers in ESP.

There were only 40 people in the group with the interesting result, so if most half of those tossed were in that group, and they'd have recorded very low belief scores, the results would have been significantly different. (I don't know enough stats to know exactly how the confidence would have changed, but the difference between the belief value of that group and the others would have been cut by about a third.)

That's a good point about who they rejected. Because it's only 40 people I would say this is definitely nothing more than a pilot study which seems to suggest it could be worth while to move on to a bigger scale.

edit: But then again I think you could make a case for throwing them out. I think the best thing would be to show the results with them included and omitted.

I think this is a great point. People brought in might have thought this test was a gullibility test and skewed their answers. It is very out of the ordinary to anyone with any founding in science to hear that science is now recognizing ESP. The question is how do you do this experiment without that possibly being a factor? What do you think of using something more scientifically plausible like life on other planets?

This is in reference to participants who did not believe that the scientists were interested in studying how individuals respond to evidence of ESP. These would be people who came into the study and thought the study was a front for another experiment. As a result, they threw out their data because they believed their responses were predicated on the belief they were being manipulated and not what was presented in the experiment.

Very small study and I would want to see the numbers before considering the implications.

Maybe they also used a suboptimal problem. A better way might be one where the vast majority of people would consider the possibility at least realistic, even if unlikely, so that the possibility of being aware of being tricked is minimised. For an example, they could say that uhh I don't know, Mongolian nomads use the behaviour of some species of bugs to forecast the weather.

With the current problem, a handful of idiots who sense scientists as the evil guys who take fun out of everything could sway the test because many people are aware that guessing cards is bogus to begin with and thus immune to the information given by the test. For an example to me the claim of the test would absolutely not matter because I would be convinced that guessing card is bogus to begin with. So the question posed in the test could not influence my answer while it could influence an idiot who is paranoid of scientists. On the other hand with a problem like the weather bugs, my answer would be 100% dependent on the info that the test gave me, because for all I know, it's at least perfectly possible, even if I'm very sceptical of the idea.

Maybe some sort of new pseudoscience would be better.