I said:
>> “Not exactly true. Science and Engineering organizations are trying to create programs aimed at both girls and boys. ”

JediBear said:
> Actually, that would make it exactly true. A program aimed at boys AND girls is aimed not at boys but at children generally.

Oh, so you meant “programs aimed at boys exclusively”. But I am confused. Why would boys need programs aimed at them exclusively? Of course, one could also argue that traditional programs tend to be male focused anyway, which is one reason why girls need programs intentionally tailored to interest them.

>>“Wait, now you are using one anecdote to refute a general trend.”

> Not so. I did not speak to a general trend, but to an absolute statement. An anecdote can be used to refute such a statement, as only one example is needed to render the statement false.

Your original remark was, “As an aside, who says women don’t go far in the sciences?” It is a recognized trend that science and engineering fields are male dominated by the numbers. In fact, the original comment by Siduri was, “Any of you who wonder why women don’t go so far in the sciences…” That comment is addressed to a trend, not any individual cases. Ergo, your response is to refute a trend by a single case.

>> “On the other hand, if her experiences are taken to be typical,”

> My point was precisely that her experience cannot be taken to be typical.

Which experiences cannot be taken as typical? Her experiences of success in the science field, or her experiences confronting attitudes of bias and expectation? While she is a unique individual, and any particular events are unique to her, it is certainly reasonable to conclude that if she received a high percentage of attitudes addressed to her being a woman, that other women might confront that attitude as well.

> My point was illustrating an alarming double-standard. If boys are outperforming girls, it’s often seen as bad. If girls are outperforming boys, it’s often seen as good. … but if we would consider female academic underperformance to be a bad thing, we MUST also consider male academic underperformance to be a bad thing on the same basis. It’s not panic, it’s just not sex-biased.

I agree that academic underperformance from any particular group is equally bad. I guess the issue is how to tell when a particular group is underperforming.

>> “Women tend to have to work harder for equivalent recognition of success. At least that is the premise. ”

> My point was that that premise is not borne out by the evidence and cannot therefore be considered true. If girls did have to work harder to succeed, they would suceed less often. Since the second is manifestly untrue, the first cannot be considered to be true.

I think you will find people arguing that the lack of females in the sciences is an indicator of less success. How is success measured?

> What’s not feminine about football?

Large bodies slamming into each other at high speeds with lots of force.

> The term “girly” implies and is defined by the stereotype of what a girl should be. In this stereotype, women don’t do math (or physics, or politics) and don’t play football (or climb trees. Or get in fights.) Doing either would be failing to adhere to the stereotype and make one “not girly.”

Okay, I guess that fits how you’re using the term. I think penny was using the term to mean “wears dresses and likes pink and purple and plays with dolls and wears makeup and jewelry”.

> This would not create a bias against girls, but only against those (whether male or female) who were meaningfully ignorant of the sport in question. The female stereotype, not the test, would be creating any such bias.

It is true the bias would be against people not familiar with the sport. Because of cultural trends where females are less interested in sports than males, this becomes a de facto bias against females. It’s not an intentional bias against women (or against anyone, for that matter), but it has the result of being a bias against women because of cultural patterns.

penny said:
> “One can do the same thing by asking boys questions about dress darts, french braiding, clairified butter etc.”

JediBear said:
> Just curious, do these have unique mathematical properties? Sports scores do.

Whether or not there are unique mathematical properties, there is an advantage to familiarity with the topic. It makes it easier to conceptualize, and shortcuts understanding when you already are familiar with the mathematical relationships and calculations. If you already regularly sum football scoring and track points, it is much easier to pick up, say, totalling the score of two TD’s, a field goal, and a 2point conversion. Someone familiar will be more comfortable running through the numbers than someone who has never heard of football, even if all the point awards are spelled out. I don’t know about dress darts and clairified butter, but french braiding talks about patterns of overlaying strands of hair. A similar topic might be knitting, where different patterns are formed by different counts on the knots.

Penny said:
> “Icelandic Girls outperform Icelandic Boys on their version of a similar test.
So, it is clearly cultural.”

JediBear said:
> Well, no. There are numerous other differences between America and Iceland (population, climate, policy, etc.,) any of which might ultimately explain the difference. Cultural differences are certainly a favored explanation, but the matter is by no means proven.

Uh, how do population, climate, and policy have effects on test performance other than cultural effects?

That clarifies the nature of your claim somewhat. Usually, the term “girly girl” does not refer to the expression of secondary female sex characteristics.

Thus my confusion. There IS a known link between secondary female sex characteristics and estrogen levels.

“These claims are also based on very small studies of
girls with MEDICAL syndromes ”

Which are necessary to produce such a condition (low-estrogen girls, low-androgen boys) in humans, since it’s not an acceptable practice in current behavioral studies to take otherwise normal children and remove their gonads.

“Look up the definition of “independent random variables”. Very few real world random variables are independent.”

I am familiar with the term. I looked it up, and it meant exactly what I thought it did. That being the case, you’re still begging the question.

“There were four sports score questions on my SAT. ”

Out of a total of…?

Hm. That’s remarkable. Of course, it’s not exactly meaningful. Anecdotes, incomplete data, and all that.

“I didn’t claim the estrogen connection was proved.”

You claimed there was a connection between estrogen levels and being a “girly girl.” Of course, you were using the latter term incorrectly, and the claim is therefore actually (and trivially) true. I’m still going to claim that your estrogen levels and baseline performance are not sufficiently established to lend any level of legiitimacy to your anecdote.

“Small studies using invalid statistical methods are just as worthless as any anecdote.”

Not quite true. Small studies using invalid statisitical methods (a claim which is still passing here without evidence) are at minimum just as worthwhile as any anecdote. Like an anecdote, they can still suggest something that is worthy of further study.

More importantly, a study being invalid doesn’t make it wrong, and no number of anecdotes opposed to it can be taken to be right, which was more or less my point.

““Anecdote” refers to a story. ”

Without regard to the truthfulness or accuracy of that story. A single incident, without context, is an anecdote and therefore statistically and scientifically meaningless.

Now, I did make an incorrect claim. An anecdote is not sufficient to provide a counterexample, only necessary. It would be better to say that a single well-documented example (such as yours of women at the top of their mathematical field) is sufficient to refute an absolute claim to the contrary.

I’m a little confused, however, as to how we even got to be discussing this. The anecdote I claimed was meaningless was your story of your sexist science teacher. There are no names or facts. The matter as presented cannot even be investigated regarding its truthfulness. Therefore, as a skeptic, I should treat it in exactly the way I would if you claimed that you had a science teacher who could fly.

“You completely misunderstood my point about “short exam”. The point is that a nearly perfect score on short exam can be ruined by a VERY SMALL NUMBER of wrong answers, so that the situation is very sensitive to bias. ”

I didn’t misunderstand your point. It wasn’t sufficiently explained, and I left with NO understanding of what you were driving at.

As it is, your point did not address my question. While it may suggest a sensitivity to bias, it does not demonstrate the possibility of sex bias.

“You also misunderstood my points about questions based on sewing dresses etc. That is: Familiarity with terms matters.”

In math, familiarity with terms only matters when they have a numerical import. Thus my question.

Seriously, how would you design a math question to bias it against even the male stereotype, to say nothing of a gender-normalized boy?

“My point about tails of distributions was lost on you, I see.”

It wasn’t exactly lost on me, I just had no idea where you were going with it. Now that it’s explained a little better, I can see your point. On the other hand, I’m still not sure how it relates to any of my points. Care to explain?

On the other hand, I’m not a statistician, and I’ll admit that you’re starting to get a bit over my head.

“The rest of your twists and illogic and your sarcasm are so great that I just give up. ”

Illogic? Where have I been illogical? Can you point out a fallacy, even one?

Sarcasm, I’ll admit to. I’m a sarcastic bastard. Ask anyone.

“I don’t have enough androgen to fight on.”

There you go making a behavioral claim based on sex hormones again.

OK, I am now really off this THREAD.

Look up the definition of “independent random variables”. Very few real world random variables are independent.

OFF THIS THREAD

There were four sports score questions on my SAT. Luckily, I had
a boyfriend who was sports nuts and I understood the questions.
A year earlier, I would have NOT, as I loathed sports in all forms.

I didn’t claim the estrogen connection was proved. In fact, I claimed the opposite. It was the Sciam Article that claimed it.

Small studies using invalid statistical methods are just as worthless as any anecdote.

“Anecdote” refers to a story. That is, medical doctors in clinical practice who would say: ” I had a patient ( no name or data given) who responded to this treatment. It doesn’t refer to the completely documented fact that certain women mathematicians have won Macarthur
fellowships for their first rate math research. Look up the word
“anecdote” in a dictionary.

You completely misunderstood my point about “short exam”. The point
is that a nearly perfect score on short exam can be ruined by a VERY SMALL NUMBER of wrong answers, so that the situation is very sensitive to bias.

You also misunderstood my points about questions based on sewing dresses etc. That is: Familiarity with terms matters.

My point about tails of distributions was lost on you, I see. Even if something is actually a gaussian bell, it is well known that the tails are not RELIABLE statistical estimates. That is why, for example MENSA, uses a two sigma cutoff–data beyond two sigma are not reliable. The
data for SAT scores used to claim male superiority at the high end are
in fact beyond two sigma. Hence, NOT STATISTICALLY RELIABLE.

The rest of your twists and illogic and your sarcasm are so great that I just give up.

I don’t have enough androgen to fight on.

“Not exactly true. Science and Engineering organizations are trying to create programs aimed at both girls and boys. ”

Actually, that would make it exactly true. A program aimed at boys AND girls is aimed not at boys but at children generally.

There are programs aimed at girls. There are not programs aimed at boys, nor are they expected.

“Wait, now you are using one anecdote to refute a general trend.”

Not so. I did not speak to a general trend, but to an absolute statement. An anecdote can be used to refute such a statement, as only one example is needed to render the statement false.

“On the other hand, if her experiences are taken to be typical,”

My point was precisely that her experience cannot be taken to be typical.

“The one you mirrored when you jokingly said “Now that women have swept the competition, perhaps we should start worrying that MEN are being excluded from math and science education.” There are people seriously concerned about male academic “underperformance”. ”

My point was illustrating an alarming double-standard. If boys are outperforming girls, it’s often seen as bad. If girls are outperforming boys, it’s often seen as good. I don’t care about the relative performance of any two arbitrary groups (boys vs girls, rich vs poor, black vs white, blue-eyed vs brown-eyed, names starting with A-M vs names starting with N-Z,) in anything. but if we would consider female academic underperformance to be a bad thing, we MUST also consider male academic underperformance to be a bad thing on the same basis. It’s not panic, it’s just not sex-biased.

“Women tend to have to work harder for equivalent recognition of success. At least that is the premise. ”

My point was that that premise is not borne out by the evidence and cannot therefore be considered true. If girls did have to work harder to succeed, they would suceed less often. Since the second is manifestly untrue, the first cannot be considered to be true.

“How so? What is not feminine about math?”

What’s not feminine about football?

“Physical rough and tumble is one thing, mental prowess another.”

The belief that women can’t be “rough and tough” is every bit as wrongheaded as the suggestion that they can’t be intelligent. Even considering differences in normal physical structure, nothing prevents girls from excelling at football.

” Are you falling for the stereotype that women don’t do math? That women can’t be smart? Are you using the term “girly girl” differently than penny?”

The term “girly” implies and is defined by the stereotype of what a girl should be. In this stereotype, women don’t do math (or physics, or politics) and don’t play football (or climb trees. Or get in fights.) Doing either would be failing to adhere to the stereotype and make one “not girly.”

Also, Penny suggested a connection to estrogen supply, which was actually the point of contention, and which I still contend remains unproven.

penny-
“It is very easy to bias a math test. First, the SAT is fairly short test, shorter yet–because some proportion of the questions are not actually used for the grade, but used to create the next year’s test–based on the result.”

Which is meaningless, unless we presume that there exists some gender difference between men and women in the ability to deal with a short exam.

“This means that the high end is determined by very few questions. All it takes is to make the context of those questions more familar to boys than girls–sports score questions work–because many girls–even now, are not interested in sports scores and have limited familarity with sports situations and terms.”

Not that it matters (see next two paragraphs) but how many such questions did you contend with in your SAT?

This would not create a bias against girls, but only against those (whether male or female) who were meaningfully ignorant of the sport in question. The female stereotype, not the test, would be creating any such bias. Even so, such questions could be, should be, and are avoided. After all, they would be testing general (not mathematical) knowledge, which isn’t the point of the test.

Unless you’re suggesting that some sexual dimorphism makes it harder for girls than boys to understand football scoring.

“One can do the same thing by asking boys questions about dress darts, french braiding, clairified butter etc.”

Just curious, do these have unique mathematical properties? Sports scores do.

“The fact that girls outperform boys in actual advanced math classes–classes that measure abstract thinking and geometric spacial skills, is far more significant that the scores at the high end of a single test.”

In determining WHAT, exactly? What’s the point you’re now trying to make? I ask because I’m lost. (Perhaps, being a man, I’m simply not as smart as you…)

Also, it should be noted that one’s actual marks in a class are usually deliberately somewhat independent from one’s level of skill in the subject of the class.

“Moreover, any mathematical statistician will tell you that the far tails of a bell distribution are statistically unreliable and have NO significance for prediction.”

Again, your point is…?

Did the study in question at some point rely on the predictive value of the far tails of a bell distribution?

If you’re talking about the Infamous Taff Bell Curve, well, that was kinda the POINT of the thing.

“Icelandic Girls outperform Icelandic Boys on their version of a similar test.
So, it is clearly cultural.”

Well, no. There are numerous other differences between America and Iceland (population, climate, policy, etc.,) any of which might ultimately explain the difference. Cultural differences are certainly a favored explanation, but the matter is by no means proven.

Also, WHAT is cultural? Female underperformance on standardized tests? Can you show me where I claimed it wasn’t?

“Al said: ” I prefer even small studies to anecdotal evidence”.”

Misattribution and misquoting in one go. You should be ashamed of yourself.

Actually, I (JediBear, not Al) said “Anecdotal evidence is inferior to even small studies,” which is true, since anecdotal evidence is statistically meaningless, and are therefore the minimum value which a study may approach.

Poor Al never used the word “anecdotal,” and did not comment on the study in question.

“I also find it interesting that when I cite the existence of recognized world level female mathematicians ( with names), it is called by Al “anecdotal”.
That is a complete misunderstanding of what that term means.”

Again, misattribution and this time a complete failure to understand the claim in question. I claimed that your experience with the sadly misinformed and logically deficient teacher was anecdotal and therefore statistically meaningless. It is. The success of the women in question is also anecdotal, but sufficient as such to refute his claim, with which I do not agree.

“What it says is that if you actually have independent random variables ( certainly NOT the case in intelligence studies)”

Isn’t it? If not, why not?