The stanford experiment had 24 participants, that's a perfectly acceptable sample size. As a matter of fact, it's a rather large one if you look at the averages I see here.
From a sample size of about 15 statistics can be used to determine how likely the result are thought to be attributable to chance.
The size was alright, but I'm talking about the fact that they only used men of a certain age group and certain race and certain class and deem the results as human nature.
Yeah, I've always questioned that practice. But the Stanford prison experiment isn't the only one that uses primarily students and males. A lot studies do at least one, most studies do both. I do wonder how much of a difference an accurate sample of the population would make, but something like that is incredibly difficult to put in practice and I don't know of a single study that has done it.
.
Does someone else know of a study that used an accurate sample of the population?
I should say I haven't seen a psychological study that uses an accurate sample of the population. (I'm a psy major, I'm not very familiar with say medical studies and their sampling habits)
But if anyone knows of a psy experiment that uses an accurate sampling of the population (preferably one that has been repeated in the more common student partipants set-up) please post!
I'm very interested in knowing if, and if so, how much of a difference the sample makes when drawing conclusions about psychological processes.
Experiments like the Stanford prison experiment are illegal in America, but if there's something like it in a different country (and if the country's government doesn't care that's it's happening), researchers can make a case to go and observe it.
Yeah, we couldn't re-create the Stanford experiments anymore, but a lot of other behavioral psychological studies are stil being done with people today. They do have very strict ethics to follow and high debriefing standards, especially if some form of deceit was present. BUt usually they have experiments that are not harmful to the participants.
Like there was one study (which was in the field of psychoneuroimmunology) where they tested the affects of anger on the immune system as measured by lymphocytes (antigens like killer t-cells). They had people play a computergame competitively against, what they thought, was another participant but actually was part of the experiment. The videogame was rigged and could only be won by the accomplice.
They predicted that people with Type A personality would have a higher reactivity (increase in bloodpressure, heartrate etc.) than Type B personalities, and they predicted that if they measured lymphocytes (part of the immune system) they would find a decreased activity in those with higher reactivity (which they presummed to be Type A's). If you're interested, they were right. Immune system response is affected by emotions and particularly by anger and hostility (2 emotions commonly observed in Type A's).
.
These types of psy studies are still being done plenty. But there's ofcourse a huge difference between a study like the one I described and the Stanford Experiment.
Most of these experiments are still being done primarily on students, and if they are measuring physical variables, then they often use only males as well (because otherwise they have to keep track of the menstrual cycle and control for any influence it may have) So all of these studies have a sample bias and I would still like to re-create one of them to see if there are significant effects from using a rather homogeneous group to determine base psychological processes...
What I can just not get over is that people use friggin correlations to 'prove' causal effect.
Scientist that do that should friggin loose their doctorates.
From a sample size of about 15 statistics can be used to determine how likely the result are thought to be attributable to chance.
.
Does someone else know of a study that used an accurate sample of the population?
But if anyone knows of a psy experiment that uses an accurate sampling of the population (preferably one that has been repeated in the more common student partipants set-up) please post!
I'm very interested in knowing if, and if so, how much of a difference the sample makes when drawing conclusions about psychological processes.
Like there was one study (which was in the field of psychoneuroimmunology) where they tested the affects of anger on the immune system as measured by lymphocytes (antigens like killer t-cells). They had people play a computergame competitively against, what they thought, was another participant but actually was part of the experiment. The videogame was rigged and could only be won by the accomplice.
.
These types of psy studies are still being done plenty. But there's ofcourse a huge difference between a study like the one I described and the Stanford Experiment.
Scientist that do that should friggin loose their doctorates.