Significant is something of a weasel word there. You haven't read the study, and you don't know what the variance is, so you're just saying it's not significant and praying to the god of PC that I haven't read the study either. I have read the study, and though I'm not a statistician and therefore not in a position to say if any particular variance is significant in and of itself, it clearly is significant that the variance was always greater for males, and showed no particular trend. The maximum ratio of male variance to female variance was in the 8th grade, where the ratio was 1.21. It was lower among 9th graders, went up in the 10th grade, and down again in the 11th.

The authors of the study (the Hyde study, the study referenced in the OP of this thread) feel that this difference is not significant enough to explain the disparity of PhDs, but that's not the same thing as being not significant at all. Indeed, the authors say that their data predict twice as many males than females should be in PhD programs. Their findings predict 33% female participation, but reality shows only 15% participation.

A prediction of twice as many males than females is rather significant.

So once again Ken, you're wrong, and not on a matter of opinion but of fact. After the beating you've taken from Drunk Vegan earlier in this thread (starting around page 14 he referenced numerous scientific studies, but you have posted not one single shred of scientific evidence), I would think you'd be more careful about what you say. At what point do you step back and say, "wait a minute, some of these things are just matters of opinion, but some things, like whether a study showed variance (where I quoted you above), are not matters of opinion - and it seems that I get a lot of those kinds of things wrong, hmm, maybe I'm wrong"

You haven't posted any evidence at all to back up your claims. parejkoj helped you out by posting one study, but you've done nothing at all. Honestly, you're not very good at this. Also, I'd be willing to bet that you didn't read parejkoj's study either, because you're probably thinking it contradicts the two studies above. Unfortunately for you, that's not the case. parejkoj's study, which can be found here concluded that the gap in average scores is disappearing, but what did it say about variance? Well, here's a screenshot from page 18 of that study:



Whatayaknow, variance. Thailand, Iceland, the UK, and Indonesia are outliers. I wonder if they've achieved this by discouraging boys to pursue math - hey, maybe we can do the same in the US! (as an interesting side note, parejkoj's study found that the gender gap in average (not variance, but average) geometry was unaffected by social equality - I guess that'll be the next thread where people like Ken G take a haughty moral stand against the facts)

Finally, just to beat the horse entirely into the ground, here's another study (we'll call this one Hedges) that shows the same thing, only more significantly than Hyde.

Let's recap the central question of this thread: what evidence exists that there is any innate difference in mathematical ability? The answer is, there a consistent difference in variance across ages and cultures. This answer is supported by (among others):
1. The Hyde study (which found this difference, in spite of flaws that would mask such a difference)
2. The PISA data (shows the difference across all cultures)
3. The Hedges study

This issue is closed. The science is quite clear. If you're unable to accept the science, I will join you in the ATM forum and patiently continue explaining it to you.

...because the author's of the study said it wasn't the best possible test. Good lord, this is like shooting fish in a barrel. Here's a direct quote from the study that you can't be bothered to read: "We computed the percentage of items at levels 3 or 4 for each state for each grade, .... The results were disappointing. For most states and most grade levels, none of the items were at levels 3 or 4. Therefore, it was impossible to determine whether there was a gender difference in performance at levels 3 and 4. The dearth of level-3 or level-4 items in state assessments has an additional serious consequence. With the increased emphasis on testing associated with NCLB, more teachers are gearing their instruction to the test."

I tried to give you an example of the effect this would have on the study, and your response was that the example was "an absurd argument." Maybe if you brought more to this debate than just your religious-like conviction that your opinion must be right, because your opinion is the only one that's moral and ethical, you might spend some time thinking about what I'm telling you. But I'm very patient, so I'll repeat myself. The Hyde study used tests that were too easy. If I gave you and Isaac Newton a test that included only questions like, "what is 2+2" then both of you would do very well on that test and I might conclude that you both have equal aptitude. But that's not true - my test is faulty. To find the truth, to differentiate between you and Newton, would require, at the very least, a test of basic logic. With that more difficult test, Newton would come out ahead.

Multiple studies, posted above, show a difference in variance. But if you use too easy a test, and if the teachers teach to that test so that the students simply regurgitate answers, then you mask the difference. And the Hyde study's authors state that the test is too easy.

I'll wait about 24 hours, and then explain all of this to you again.