Quote:
Originally posted by Aelia Jusa:
Actually, people are merely being good scientists. Anyone who accepts faithfully a news report of a scientific study uncritically is being rather foolish, and very unscientific. You don't necessarily need counter-evidence to reasonably hold the position that this report of this study will not be taken as factual without more evidence - that's not ignorance, that's science. There is no link in the article to the actual study, as it was written up in a journal, and there are no details whatsoever of the design in the article, so we can make absolutely no claims about its accuracy. In order to do that, we need to know who the participants were, how many there were, when they were tested, how they were tested, how many times they were tested, what they were tested with, what the reliability and validity of the test instrument was, what controls were used, what the actual difference in IQ scores found was, what the significance level was, what the confidence interval for the means for each group was, what the standard error was, what journal was it published in, and so on and so forth. Even in light of all that evidence, a good scientist knows that in order to make a strong case, you need replication. You can't even get a foot in the door of one of the good journals these days without a replication built into your study, let alone someone else doing one in a different study.
|
However, I can't say I remember anyone who dismissed this study saying any of those things. if they had, that might have been valid... However, that wasn't what was given as a reason. What was given as a reason was logic that, from what I can tell, didn't work, or just plain out-of-hand dismissal.
Quote:
Originally posted by Aelia Jusa:
Actually it doesn't mean that at all (necessarily). Without knowing the distributions of IQ scores for men and women, we can't make any sort of claim about how many men and women have high and low IQs on the basis of the mean for each group.
|
Hmm... what exactly do you mean? Are you saying that it may be that men are more closely grouped around the average, while women have more variance, but a slightly greater number at the low end as well? Hmm... hadn't thought of that, actually. Good point.
Quote:
Originally posted by Aelia Jusa:
Especially for IQ tests were you can only really give a range of scores your IQ would fall into given the reasonably variable test-retest reliability, a difference of a couple of points between two groups is probably a meaningless, despite statistically sigificant, difference.
|
Meaningless? Sure. But it's still there, assuming the results weren't based on an error. [img]tongue.gif[/img]
[ 10-05-2003, 08:35 PM: Message edited by: Encard ]