Study on Skin Cream and Gun Control (Title is slightly misleading)

Originally posted by Oliver North
on the study you posted, maybe it is my background, but I'm not sure what the surprise is suppose to be? People's beliefs and expectations impact how they perceive incoming information. In fact, in almost all cases it is hugely beneficial that they do so, as the opposite would be behaving in a world where previous experience had no impact on your current behaviour, meaning organisms wouldn't learn.

Seems like they're looking specifically to deal with the argument that people are being influenced mainly by simple ignorance, ie the general public doesn't know how to think about the problems at all so the influence of political beliefs might be very small. What they found is that even people who are good at math (which they show matches with performance on the neutral version of the test) are very poor at the political version of the test.

Originally posted by Oliver North
on the study you posted, maybe it is my background, but I'm not sure what the surprise is suppose to be? People's beliefs and expectations impact how they perceive incoming information. In fact, in almost all cases it is hugely beneficial that they do so, as the opposite would be behaving in a world where previous experience had no impact on your current behaviour, meaning organisms wouldn't learn.

I think the study was checking to see if humans would think for themselves rather than using their bias to make an assessment of a situation. Is this called "critical thinking" in your field of study?

IMO, there is nothing natural about comparing 4 sets of data. I don't think "bias" should play a part when critically thinking (please, correct my use if it is wrong) about this data when they run across it. The default should not be to ignore that data.

However, one could argue that selecting the response that best fits that person's idea of "how it should be" is also a measure of what evolution does for default human decision. Selecting the one that does not fit with their "social group" could result in group exclusion.

Would you say that it is possible that selecting the correct answer (when it specifically runs against their political agenda) shows that the person is either ignoring their instinctual behaviors or that they have a new ability (this may be a false dilemma fallacy).

Originally posted by Symmetric Chaos
Seems like they're looking specifically to deal with the argument that people are being influenced mainly by simple ignorance, ie the general public doesn't know how to think about the problems at all so the influence of political beliefs might be very small. What they found is that even people who are good at math (which they show matches with performance on the neutral version of the test) are very poor at the political version of the test.

ah, alright, that makes more sense

Originally posted by dadudemon
I think the study was checking to see if humans would think for themselves rather than using their bias to make an assessment of a situation. Is this called "critical thinking" in your field of study?

how do you think for yourself without being biased by yourself?

that is sort of my point... there is no neutral "self" detached from your past experience that can think about things

Originally posted by dadudemon
IMO, there is nothing natural about comparing 4 sets of data. I don't think "bias" should play a part when critically thinking (please, correct my use if it is wrong) about this data when they run across it. The default should not be to ignore that data.

"bias" is also what allows you to know what the numbers and words you read mean... I'm sympathetic with the idea that people should try to let data speak for itself, but in this case, it took three of us, who are fairly knowledgeable about stats, a page worth of discussion to even figure out the right way to analyze the data, and we got several results depending on what we set alpha to or whether we used one or two tailed tests.

math, especially stuff relating to statistics, is something humans are terrible at, even people with advanced math degrees. for all the people who can calculate the root of 238474593457 in seconds in their head, there aren't many savants of probability.

Originally posted by dadudemon
However, one could argue that selecting the response that best fits that person's idea of "how it should be" is also a measure of what evolution does for default human decision. Selecting the one that does not fit with their "social group" could result in group exclusion.

I don't think they were suggesting people were trying to fit in...

actually, you just fell victim to the very type of "critical thinking bias" you bemoaned above. 😉

if anything, it would be closer to cognitive dissonance

Originally posted by dadudemon
Would you say that it is possible that selecting the correct answer (when it specifically runs against their political agenda) shows that the person is either ignoring their instinctual behaviors or that they have a new ability (this may be a false dilemma fallacy).

no

Originally posted by Oliver North
how do you think for yourself without being biased by yourself?

that is sort of my point... there is no neutral "self" detached from your past experience that can think about things

I agree but it is a weird philosophical degree that you're describing...I think.

Basically, one should just do the math and leave it at that. Not all of them fell victim to adding their own little flavor to the answer. However, I am betting that (like I suggested earlier) if they selected people for this study specifically because of their stances on gun control, they would get much more polarizing results. Not all left wingers are against guns (some are very much for guns). Same with the right wingers: not all are for guns.

So the people that are good with math that did not show the measured bias on the test (the H1), it is possible that a significant portion were just not polarized enough on the topic to measure them. More directly, I think they could get a stronger “in favor of H1” result.

Originally posted by Oliver North
"bias" is also what allows you to know what the numbers and words you read mean...

I think "bias" is so varied, even from person to person, that lumping the political bias they were measuring in with the visual processing of information (how our brain quickly processes symbols so we don't have to think about the whole dang symbol, each and every time we view it) is very different.

Originally posted by Oliver North
"I'm sympathetic with the idea that people should try to let data speak for itself, but in this case, it took three of us, who are fairly knowledgeable about stats, a page worth of discussion to even figure out the right way to analyze the data, and we got several results depending on what we set alpha to or whether we used one or two tailed tests.

Well, what we were doing was more for fun that what was actually necessary to correctly answer the question in the study. All we had to do was find the percentage (check the youtube video) and leave it at that. Anything above and beyond that is just for funsies.

Originally posted by Oliver North
math, especially stuff relating to statistics, is something humans are terrible at, even people with advanced math degrees. for all the people who can calculate the root of 238474593457 in seconds in their head, there aren't many savants of probability..

1. There is no root to 238474593457. I did not calculating in my head or on a calculator. It is an intuitive guess. A number that is odd, has a 1's digit that is a prime number that is greater than 5 (only 7 fits into that category), is far more likely to not have a square root than not. I am so confident that I won't even bother checking. estahuh
2. I do not know of any probability savants.

Originally posted by Oliver North
I don't think they were suggesting people were trying to fit in...

I know: I am suggesting that. I suggest that we (actually, they) are really just measuring a "fit in" behavior of humans.

Originally posted by Oliver North
actually, you just fell victim to the very type of "critical thinking bias" you bemoaned above. 😉

I sure hope not. The problem is, I know so little regarding that topic that I just have to take your word for it.

Originally posted by Oliver North
if anything, it would be closer to cognitive dissonance

Yes, that makes far more sense, now that I think about it.

Originally posted by Oliver North
no

How come some get it right despite their political leanings? How come some, who are clearly skilled in math, getting it wrong (they clearly know what they are supposed to do but make a conscious and deliberate effort to not do it)?

I love speculating. 🙂

I imagine the issue is, when you ask people a question of a political nature, they no longer look at it in strictly numeric terms. They don't think it is a math question any more.

why some people do or dont, idk

For pooled sample variance, where you assume that there is common variance between the two samples, we have

S_p^2 = (n_1 s_1^2+n_2 s_2^2)/(n_1+n_2-2)

that is, sample pooled variance is equal to the size of sample 1 times the st dev of sample one plus the size of sample 2 times the st dev of sample 2 all over the sum of samples 1 and 2, minus 2

Wolfram alpha code:

http://www.wolframalpha.com/input/?i=%28n_1*s_1%5E2+%2B+n_2*s_2%5E2+%29+%2F+%28n_1+%2B+n_2+-+2%29

There are better versions of this test, but this is all i have time to fill in for now

I could not find my master spreadsheets at home. I have no clue where they are. But, basically, it takes you through all the formulas (has examples and tables) you would need to be a researcher. I know I gave it to some students. 🙁

It was probably on my old work PC (changed jobs...old PC had lots of good stuff). Had about 20 different worksheets with multiple formulas, examples, and sometimes images. It was many week's of effort. I am so saddened by this that I may end up sulking. 🙁