Interesting series of documentaries about feminism - Page 32

On April 10 2014 06:33 Darkwhite wrote:

If it was just about representation, then why does it matter whether female characters are having a conversation, and whether they are conversing about men? Why not just count female characters or female screen time and be done with it?

On April 10 2014 06:39 ComaDose wrote:

Adding those stipulations better highlights the differences between male roles and female roles in movies.

EDIT: being "represented" or w.e. isn't just being on screen and talking.

On April 10 2014 06:42 Darkwhite wrote:

How? Why?

What more is there to it, that the Bechdel test is supposed to gauge?

On April 10 2014 06:46 ComaDose wrote:

well we see that many more men talk to people of their gender about things that aren't women, than the opposite in movies for an obvious one. This means they are talking about problems that are not romantic etc.

I thought your quote was perfect thats why i edited it in late when i saw it "the depth of their stories, and the range of their concerns"

On April 10 2014 06:59 Darkwhite wrote:

Is Gravity failing and Transformers passing the test then examples of false positives and negatives, or would you say Transformers better represents females and deal with their full range of concerns?

This doesn't mean that a movie that passes the test or not is more or less sexist than another movie. This "test" applied to one movie doesn't say anything about it

On April 09 2014 23:58 Darkwhite wrote:

Here is a list of the current top 47 chess players. Note that all of the following 23 countries are all represented: Norway, Armenia, India, Russia, Italy, Bulgaria, the United States, Azerbaijan, Cuba, France, Israel, Ukraine, England, the Netherlands, Hungary, China, the Philippines, Poland, Vietnam, Georgia, Czech Republic, Germany and Latvia.

Somehow, all of these countries are represented, despite the vast differences in culture, prosperity and interest in chess between them. Somehow, cultural effects are powerless to prevent this astonishing diversity. Yet, somehow, the best women currently occupy ranks 58 and 176. One might think the cultural barriers to becoming a world class chess player would be more formidable for some kid from the Phillipines than a woman right in the midst of the chess paradise that is Eastern Europe.

On April 10 2014 11:06 Yoav wrote:
The Bechdel test is only useful in aggregate. The point is that far more movies fail it than the reverse equivalent. It's just a very objective way of pointing out that our media culture is overwhelmingly dominated by men and male concerns. Going with it on a case by case basis is as silly as saying that all movies with black bad guys are racist, because a society where black people were villainous the majority of times they were portrayed WOULD be racist.

The Bechdel test tests society, not individual movies.

On April 10 2014 07:40 ComaDose wrote:

Oh okay. Yeah like i said if the Bechdel test is used as a pass/fail for representing women then it is heavily flawed for sure, i know some people might use it that way even though it wasn't really meant to be used at all. In the context of the original point it stands useful when applied to all movies as a means to quantify representation of women because the odds of "false positives" or negatives is not higher then relevant results so its not a paradox. At the very least it shows the amount of movies that don't have two named women, how rarely women talk to each other in movies and how often when they do talk its about men. Comparing this to the opposite shows a significant margin and serves to cancel out the "false positives" which is an interesting point.

Regardless it shouldn't be necessary to use a test to observe the problems with the way women are on average represented in Hollywood these days.

On April 10 2014 11:06 Yoav wrote:
The Bechdel test is only useful in aggregate. The point is that far more movies fail it than the reverse equivalent. It's just a very objective way of pointing out that our media culture is overwhelmingly dominated by men and male concerns. Going with it on a case by case basis is as silly as saying that all movies with black bad guys are racist, because a society where black people were villainous the majority of times they were portrayed WOULD be racist.

The Bechdel test tests society, not individual movies.

On April 10 2014 09:46 kwizach wrote:

I invite you to read Merim Bilalić, Kieran Smallbone, Peter McLeod and Fernand Gobet, "Why are (the best) women so good at chess? Participation rates and gender differences in intellectual domains", Proceedings of the Royal Society B, 22 March 2009, vol. 276 no. 1659, pp. 1161-1165. They show, through a statistical analysis, that the crushing differences in number of players between men and women account for 96% of the observed difference in results between the two. For the remaining 4%, you can turn to the cultural factors I mentioned, notably the stereotype threat effect that I discussed with Jumperer.

On April 11 2014 01:33 Stratos_speAr wrote:


The fact that throughout the history of education women have been both passively and actively discouraged from studying or being good at math and still are today?

On April 11 2014 00:28 Darkwhite wrote:


This is the sort of nonsense you get when you assume that participation and skill are independent variables. If you apply the exact same methodology to basketball, you will similarly purport to show that tall players are overrepresented at the highest level, not because being tall is itself an advantage, but because there are more tall basketball players. What's really going on is that taller players perform better, increasing the chances that they keep playing, while coaches actively recruit and develop taller players - it's easier to make a tall player good, than a good player tall. Skill and participation are not at all independent.

On April 11 2014 00:28 Darkwhite wrote:
Here is a sample where females outnumber males, so that the participation bias should pull in the other direction. What's the non-biological explanation this time around? + Show Spoiler [img] +

On April 11 2014 03:04 kwizach wrote:

Actually, that is what happens when someone doesn't bother to read the scientific article he was presented with. Your argument is addressed on p. 1163. In short, there is no evidence that supports men have any kind of biological advantages for chess. In addition, "drop-out rates for boys and girls were similar" (see Chabris, C. F. & Glickman, M. E. 2006 "Sex differences in intellectual performance: analysis of a large cohort of competitive chess players", Psychological Science 17, pp 1040–1046). This means that the selection process is not a matter of boys performing better and therefore continuing to play more than girls, unlike your basketball example. Women simply do not turn to chess initially as much as men (for sociocultural reasons), and the ones who do perform virtually exactly as well as should be statistically expected from their numbers, the extremely small difference being explained by cultural factors.

1. As argued by many scholars who have studied the topic, it is problematic to use S.A.T. scores to evaluate gender performance in mathematics for several reasons, the most important of them being inadequate sampling. First, it is only a specific part of the general population which takes the test, preventing generalization. Second, there are more women who take the tests than men. As Janet Hyde writes, "assuming that SAT takers represent the top portion of the performance distribution, this surplus of females taking the SAT means that the female group dips farther down into the performance distribution than does the male group" (Supporting online material for the article I cite next, pp. 2-3). In fact, if you take a look at ACT scores, ACT being also a test taken by students going to college, there is no gender gap in scores in states where the test was administered to all students. See Janet S. Hyde et al., "Gender Similarities Characterize Math Performance", Science 320, 25 July 2008, pp 494 ff. and the supporting online material.
2. In the US and some other nations, there are actually no longer gender differences in mathematics performance in the general population. Some gender differences do remain in the U.S. among the most mathematically talented, but the gap has been closing steadily and it does not exist in some other nations, meaning that sociocultural and not biological factors are behind it. I direct you to Nicole M. Else-Quest et al., "Cross-National Patterns of Gender Differences in Mathematics: A Meta-Analysis", Psychological Bulletin, Vol. 136, No. 1, 2010, pp. 103-127 and Janet S. Hyde, Janet E. Mertz, "Gender, culture, and mathematics performance", PNAS, Vol. 106, No. 22, 2009, pp. 8801-8807.

On April 11 2014 04:24 Darkwhite wrote:

The statistics only account for people who have played chess in an organized environment. This is a fraction of the population, which is not selected at random nor only for sociocultural reason - primarily, they are selected for whether or not they find the game interesting and enjoyable.

The article hinges on the assumption that the women who do play chess have wound up there by some sort of accident or happenstance, and that there is just as much talent among the women who don't play. Do you really find that likely?

On April 11 2014 04:24 Darkwhite wrote:

It's interesting how SATs get criticized for inadequate sampling while chess statistics are fair game. I'm sure people with FIDE ratings are not a specific part of the general population, preventing generalization.

On April 11 2014 04:24 Darkwhite wrote:
The excess women taking the SATs actually works in their favor; the graph I linked gives the straight ratio without any normalization. Despite a larger number of women being tested, there are still significantly more men with the best scores. Notice that this larger talent pool is the exact reason given, with regards to chess, for the lack of top female players. While it is true that the excess women could hurt the female average score, it certainly shouldn't hurt their absolute representation, at any level.

On April 11 2014 04:24 Darkwhite wrote:
It is true that Hyde has sort-of pretended to show that there is no gender gap in maths. She has done this through sheer intellectual dishonesty, in particular:
- including pre-pubescent students in the sample; the same sort of group where girls are taller than boys
- using minimum-level tests such as the NCLB, which doesn't distinguish between average, above-average and exceptional performances, but primarily singles out ineptitude

On April 11 2014 04:24 Darkwhite wrote:
Even so, when Hyde has done everything in her power to rig the testing to diminish the gender gap, one problem remains - that, while averages of males and females approach each other, the variability of male scores remains significantly higher. What this means is that, when you look at only the most talented, such as the 95 percentile, you will find an overwhelming male dominance. Exactly like we see in the International Mathematical Olympiad, or in universities.

I would be interested to see exactly which countries do not have a significant gap between men and women at the higher levels; many are included here, in what seems a fairly robust trend.
+ Show Spoiler +
Arrows show that the outliers typically aren't stable from the 2003 to the 2006 PISA test.

On April 11 2014 07:15 kwizach wrote:

The article does not make any assumption on why the men and women who play chess do so. It looks at the respective chess performances of men and women in chess and demonstrates that 96% of the difference in representation in rankings can be attributed to the respective numbers of players of the two population. The population of women who play chess virtually does not play statistically worse than the population of men. For the remaining very small 4% difference, I provided evidence of the role of cultural factors in the form of scientific research done in social sciences on that very topic. Thus, chess rankings & performance simply does not support the idea of greater abilities for males.

Since you cannot dispute these numbers, your objection is that there are fewer women because women are inherently worse at chess (due to lower relevant natural abilities). This runs into four problems.

1. You would expect this to still show among the population of women that does play chess, but it doesn't. The opposite is true.
2. The proportion of women starting to play chess then stopping (and therefore not being taken into account by Bilalić et al.) is the same as the proportion of men starting to play chess then stopping (and therefore not being taken into account by Bilalić et al.), as shown in the study by Chabris and Glickman which I referred to in my previous message. This means that the difference in numbers in men and women competing is simply not explained by women being unsuccessful/worse than men and therefore dropping out without having a chance to appear in the sample, since they drop out at the same proportion as men.
3. The remaining explanation is that there are simply way less women trying/engaging in chess than men in the first place. Unless you are going to tell me the explanation is that women are more physically deterred by a game with black & white squares (which would still be irrelevant to intelligence), this means that sociocultural factors explain the difference.
4. Beyond these points, there is zero evidence of biological factors explaining any sort of difference in chess performance.

To sum up: looking at chess performance does not, in any way, support the hypothesis that men have abilities that women don't. Not a single aspect of the issue supports the hypothesis.


You seem to be confused about who is arguing what with respect to the chess population. You brought it up (after Jumperer) to argue the point that men had superior abilities than women. I explained to you that your example did not support your point since the chess performances of the two populations are statistically equivalent. I'm not trying to argue that people with FIDE ratings are not a specific part of the population, you were making that generalization, and I showed you that even among the specific population you had chosen your point was not true.


Yes, I was addressing average male and female scores. The problem I referred to with regards to the higher number of women, potentially including more less-capable candidates, still impacts the M/F ratios at the lower levels. Beyond this, however, the sampling problem remains entirely. Since the test is mainly taken by those who wish to attend college, and men are very much in the majority when it comes to studies in engineering/physics, for example, you could expect more men that have put an emphasis on maths to take the SATs than women. I'm not sure why you think SAT scores are at all representative of anything, let alone evidence of the role of biological factors.

With regards to the parallel you draw to chess, in chess women statistically performed just as well as men. In the case of SATs, they did not (with plenty of non-biological possible reasons, as I said). In addition, the possibility of a negative impact of higher numbers of women on their average scores (and on the M/F ratios at the bottom) raised by Hyde pertained to the attributes of the excess population for women. I'm not sure what comparison you're making with chess here.


The articles I cited used the tests which allow for the most accurate comparisons among the U.S. population and among other countries (TIMSS and PISA tests for cross-countries). They take into account different stages, and certainly do not present only averaged aggregates.


This objection is pretty funny considering that if you had bothered to look at the three articles I cited, you would have noticed that it is completely false: the variability hypothesis is tested each time, and extensively so in Hyde & Mertz (2009). I even referred to those results when I wrote "Some gender differences do remain in the U.S. among the most mathematically talented, but the gap has been closing steadily and it does not exist in some other nations, meaning that sociocultural and not biological factors are behind it". Several measures are examined and explained at the 95th percentile and above, exactly what you argued they should have done. Like I said, a gap does remain in the U.S., but it is smaller than before, it varies across ethnicities (for Asians, it is girls who score better above the 99th percentile), and it is nonexistent in some other countries, pointing clearly towards the role of sociocultural factors. If you read the articles, you will see that Denmark and the Netherlands are examples of countries where males did not have greater variability than females.