Solomon Asch (1907-1996) a social psychiatrist did an experiment meant to show group pressure.

The subjects where shown a plank with one line,then shown three other planks with lines of different sizes ,only one of them being equal with the first one.

The subjects were introduced in a room one by one,with other 6 people(Asch's undercover people).

The first people (asch's people) intentionally gave an obvious wrong answer and last person(the genuine test subject) was asked to answer which was the equal sized line.

70% of all subjects gave in to peer pressure and answered as Asch's people ,when the correct size line was obvious to anyone.

When subjects did this alone 99% of them gave the right answer.

When they did it in the group, affected by peer pressure ,only 30% of them answered right.

http://en.wikipedia.org/wiki/Asch_conformity_experiments

If the numbers posted are not accurate don't blame me, they're from the manual issued by professors of SNSPA (a romanian uni) i've been studying for my sociology admitance exam on saturday.
The manual is approved of the ministry of education and research here.

The number of the "sample" is not known to me.

With 1000 people sample the accuracy of the percentage is +/- 5% ,if anyone knows the "sample" size please post it.

The correct line was obvious for even a small child,the difference between the one and the others being huge.

The number is much larger than 70%.

That explains so much.

There are four lights!


I think it would be worth knowing just how different the lines were. If they were fairly close then I could see people assuming they're falling for an optical illusion or something, there are certainly plenty of optical illusions out there that make things that are the same size appear different or vice versa.

The length of a line is also fairly unimportant, I'd be curious to see the results of a similar study with something equally objective but substantially more important.

I took highschool social sciences class, too.

Yeah but this time science says it hhah

Also look at the wiki link to see the lines.

The correct one is obvious as it is very different from the other ones.

Yeah I've seen results from similar studies like that, it's kind of sad. If I was in that situation I would look at everyone and ask "What the [censored] is wrong with you guys?"

Actually, from certain angles it looks like it could be b OR c.

http://xkcd.com/610/

I don't get it. Why would they say the wrong answer knowing it was wrong just because other people did it?

*Yawn* Welcome to Psych 101, people.

And I'm just going to say it anyway... Asch's studies are of a limited use in judging people as "sheep." Of course, if you want to make the generalization based upon one study...

If it was a sample of 1000 people ,the margin of error is +/- 5 % for any population size.

Making the study representative for the entire population.

I don't know the sample size,unfortunatly.

From what i know other study's yielded similar results.

You can't get more objective that that,with lines.

I am not a sheep...Baaa

123 male students.


Which means the only thing we can be sure of is that on average, 70% of male american students are sheep.

Correction: 70% of male students half a century ago.

Any study of this kind can be a generalization if the subjects are over 1000.

The statistic laws say that no matter the population the results will be the same.

If you cook soup in a small bowl and taste it with a spoon to see if its ready,do you feel the need for a larger spoon when you taste it from a larger bowl?

baaaa

While that's usually going to be true, that's an arbitrary number to quote. You can't actually know anything about the margin of error on a statistic with just the sample size; you also need to know the variance of the values for that statistic within the sample.

It is the case, though, that social science experiments that find statistically significant results usually use something more in the realm of 30-300 subjects; if it took a sample size of 1000 to reduce your confidence intervals down to acceptable sizes, it would mean you were measuring an http://en.wikipedia.org/wiki/Effect_size so small that I can't imagine why anybody would care about your results.

*edit* One thing that's worth noting: According to the wiki link, even in the peer-pressure condition, accuracy was still 68% (i.e., people resisted peer pressure more often than not). I'm assuming that "70%" figure the OP is quoting refers to the fact that 75% of people answered at least one of a long series of questions incorrectly. That's still an interesting finding (since only 1 of 35 in the control condition answered anything wrong), but it's not the "people usually cave to peer pressure" that the OP sounds like it's espousing.

One, you haven't examined the study for biases in recruiting or sampling. A biased sample is representative, not of the population you claim to study, but of the researchers' biases, notwithstanding any argument based on sample size. turns-the-page already showed that the study was biased, and you have not refuted that.

Two, the subject is inflammatory; discuss decently and in order or it will be locked.

If the students where 123,the accuracy of the percentage is +/- 9%.

n= 1/square root of N *100

Where n is n +/- %

and N is the number of people in the sample

It may be further be distorted(the accuracy) if you want to count in the females since they where not in the experiment.

It's no surprise. For most people, all the evidence they need that something is 100% factual is that someone, somewhere, agrees with them.

Flossie! :wub:

I pretty much take back what all i said, since i started based on lack of vital info.

I redid the percentage for the 123 (people)---not males only---and that's +/- 9%.

Adding the female element further destroy's its accuracy.

Thanks turns-the-page for the info.

The females, non-americans, anybody not living in the social conditions of the 1950s....

Did they get to see all four lines together at the same time? If someone showed me the card in the Wiki link then took it away and showed me the next cards and asked the question I would not be sure of my answer at all.

It had to be random selection, 123 all male is not random selection lol

It holds no validity sadly.

My thoughts exactly. I mean, really. How do you know you won't do the same in such a situation unless you're in it?

I like to think myself more feline than a sheep. More like one of http://www.nationalforestlawblog.com/lynx.jpg, if we had to choose an animal to represent me.

But still, I assume it is possible if put in a similar situation I'd follow herd instinct. I just really hope I don't.