Thursday, June 10, 2010

The SAT Math Test: It's Rigged to Favor Boys?

From an email I received yesterday:

"Regarding the sexist Tierney article in the New York Times (featured Tuesday on CD), there's lots of evidence showing that a much larger part of boys' test scores is due to tricks. Studies show males take more risks than females, and the SAT rewards taking a risk by guessing when you don't know the answer. Thus, the SAT is rigged to favor boys this way because it rewards riskier test-takers and boys are more riskier than girls. Girls' unwillingness to risk getting a penalty for a wrong guess has been shown to significantly reduce their scores (Gender Bias in College Admissions Tests, Research by FairTest (January 14, 2009).

And of course the test favoritism shows when students actually try to apply their knowledge. That's why boys who get the same test scores perform worse than girls in college. For example, a 1995 University of CA, Berkeley study found that females with identical academic indexes as males earned higher grades in every subject including math and physical sciences. The report concluded that schools should add 140 points to the women's index because the SATs underpredicted their ability.

The SAT is rigged in many ways to favor boys. For example, the ETS did a study that found that girls' SAT scores improved dramatically when the time limit was removed but boys’ scores remained about the same. Since when the time pressure is removed, girls perform much better, the test is obviously testing things other than actual knowledge. This can be changed and the College Board is aware if it, but they have refused to make this minor change that would dramatically improve girls' scores in comparison to boys' scores. Thus, the people in charge of the test knowingly do things that favor boys."



At 6/10/2010 10:58 AM, Anonymous Anonymous said...

This seems like a pointless exercise when you look at the number of females in college, the number rec'ing degrees, etc.

My response to the author would be, "fine, we'll adjust the SATs. But we'll re-adjust until you have 50% male, 50% females getting into each school."

At 6/10/2010 11:04 AM, Blogger Bret said...

No links to studies and I'm skeptical.

For example, college math tests are also timed, aren't they? Is there more of a penalty for wrong answers on college math tests? If not, how would the boys "perform worse"? By what measure?

Nearly every time I see the phrase "studies show" and there's no link to said studies, it ends up the studies have lots of caveats and are not really applicable to the discussion at hand.

At 6/10/2010 11:05 AM, Blogger KipEsquire said...

Actually the SAT discourages risk-taking by penalizing wrong answers and by signaling question difficulty (questions are arranged from statistically easiest to statistically hardest).

Either way, candidates who dislike these attributes can opt for the ACT, which does not penalize incorrect answers and randomizes question order (and has longer section times, btw).

At 6/10/2010 11:06 AM, Anonymous Anonymous said...

I guess we should add points to males in the non math portion of the test or more absurdly allow males to fire guns to stoke their testosterone before taking test to heighten mental awareness!

At 6/10/2010 11:14 AM, Blogger juandos said...

Thank you Charles Murray: We Can't All Make the Grade

At 6/10/2010 11:16 AM, Anonymous Anonymous said...

The SAT test is a perfect measurement of how well someone can perform on the SAT test. I am not sure how well the knowledge that it takes to do well on any test transfers to usable skills in the real world.

I have students who ace a federal test given by a third party that I would not hire if I had a business. It does not matter how well someone tests on any subject if they cannot find a way to use that knowledge when they are not sitting in a chair in a controlled environment.

At 6/10/2010 11:17 AM, Blogger epv said...

Maybe the method that college courses are graded is biased towards women? That could help explain why there is such a large imbalance in the degrees being awarded to women versus men.

At 6/10/2010 11:54 AM, Anonymous Tom in Maine said...

Just to be politically correct and not hurt anyone's feeling, they need to keep changing the test so everyone scores equal and no one feels un-favored. Maybe the grading should change to take into account gender, race and sexual preference issues? (jk)

At 6/10/2010 12:56 PM, Anonymous Anonymous said...

The lack of risk follows right through college. This very site so recently compared college majors: which gender goes for the "easy" majors (a rational economic decision if one is aiming for a job in a market that rewards "a degree" versus "the major," e.g., pretty much anything unionized, esp., say, teaching).

Ironically (and an area for research), it seems that if you have a group of individuals, many of whom are making individually rational and self-interested decisions (a la Adam Smith), it appears that the outcomes (mean, mode, and outliers) will differ from a "riskier" set of individuals.

(Even there: perhaps by shooting for a very high goal, if one falls short one may still fall above s/he who shoots for the "sure thing"?)


At 6/10/2010 1:10 PM, Anonymous Anonymous said...

The fact that more women are graduating college is proof that males have a higher aptitude for math:

8 Reasons College Tuition Is Next Bubble to Burst

At 6/10/2010 1:24 PM, Blogger Aaron McNay said...

Even if we assume the information that the individual provided is correct, it does not refute, it in fact confirms, the broad point that is trying to be made, which is: women are different from men. We can discuss what these differences actually are (intelligence or levels of risk aversion), which is important. However, what causes these differences does not change the basic point that there are a large number of possible reasons why men outnumber women in science and other fields, other than discrimination.

At 6/10/2010 1:29 PM, Anonymous steve said...

You mean we can't solve the 'tricks' problem by simply advising girls to guess and "take risks' when they don't know the answers? This was what everyone was advised to do back when I took the SATs some 40 years ago.

At 6/10/2010 1:48 PM, Anonymous Roscoe said...

What does "identical academic indexes" mean? It sure doesn't sound like "identical SAT scores."

It seems odd to acknowledge gender differences in some cases, "boys are more riskier than girls," in an attempt to dispel gender differences on the SAT math scores.

When I went to Berkeley in the 80's, it appeared that women were much more responsible about going to class regularly, doing their homework, etc. Homework counted for 30% to 40% of the grade at Berkeley, which was high when I compared to friends going to other schools. It probably cost me at least half a GPA point overall since I didn't do homework very regularly or attend class so regularly, but I did do very well on exams.

Also, there were always two different levels of courses for math and science. So if you took Physics 7 it was for the hardcore math/science/engineering majors, and Physics 8 was for non-majors and was much easier. Maybe I have those backwards, but there were two tracks. Calculus had a similar split. I took the harder physics class and on the first exam scored in the 20's out of 100. Which sounds bad except the class average was in the mid-20's. Being Berkeley, someone did get a 98. Yeah, I should have taken the easier physics class.

The math/science majors had to take the harder courses as pre-reqs for other courses in their major. Everyone else could take the easier one or take the harder one. Since men were disproportionately represented in math/science/engineering, they on average had to take the harder series. So raw GPA doesn't necessarily mean anything, it would have to be viewed course by course. And would have to be further adjusted for some of the whacky science classes that counted as a science class.

It is Berkeley after all, so there were a lot of questionable classes that counted for the math/science requirement. One of my geology classes was really a sociology class where they sung the praises of primitive societies and belittled the modern western world. I'm pretty sure I didn't go to that class much after the first few sessions.

At 6/10/2010 2:40 PM, Anonymous Anonymous said...

If the claims made about gender and risk taking are true, then a test that is neutral towards risk taking would be biased against men, no?

At 6/10/2010 3:30 PM, Anonymous Anonymous said...

The tenant that SAT testing is rigged to favor boys, if so who did the "rigging" ?
I ask myself how long this has been going on and who the evildoers are
The Trilateral Commission perhaps?
Any good conspiracy theory needs names and dates, sheesh.

At 6/10/2010 3:30 PM, Anonymous Anonymous said...

The tenant that SAT testing is rigged to favor boys, if so who did the "rigging" ?
I ask myself how long this has been going on and who the evildoers are
The Trilateral Commission perhaps?
Any good conspiracy theory needs names and dates, sheesh.

At 6/10/2010 5:48 PM, Anonymous Anonymous said...

People in glass houses shouldn't throw stones. Where were these whiners when the SAT was rejiggered to favor girls by the addition of a second, hand written, verbal section? That's why the new SAT maxes at 2400 points, not 1600.

At 6/10/2010 5:51 PM, Blogger Milton Recht said...

Students who take the PSAT earlier, score higher on the SAT than those who do not take the PSAT. Also, those who take SAT prep courses tend to do better on the SAT.

Not clear if cause and effect or underlying characteristic of those who take PSAT and prep courses.

Need to see if male female ratio disparity in prep course taking and in PSAT taking can account for any of the disparity.

At 6/10/2010 7:24 PM, Anonymous Titus Pullo said...

"Boys are more riskier than girls." By saying that, isn't the author lapsing into stereotyped, bigoted speech?

At 6/10/2010 7:37 PM, Blogger Jason said...

Rigged, I doubt it.

Really college grades and SAT scores are not the same thing. The SAT has defined test criteria, as a specific test to determine a gauge of knowledge and intelligence.

College grades are an indicator of how students did in their chosen curriculum and electives. A stacked deck, so to speak. So the comparison is apples to oranges.

Better idea, compare like subjects to see how it stacks up. My guess is that then you will confirm the SAT data. But since that would not be what the liberal elite educational block would want, don't hold your breath.

At 6/10/2010 8:42 PM, Anonymous Anonymous said...

It's been a while now, but I'd swear the instructions I received were *not* to guess; that you lose more points if you guess wrong than if you leave it blank. If that's right and if boys really are bigger risk takers, then the SAT would seem to favor girls, not boys.

At 6/11/2010 5:46 AM, Blogger Unknown said...

The ability to assess risk and make good decisions quickly is critical for success in the business world. Often, timing is everything. Low risk takers (both male and female) who insist on taking lots of time to analyze all the facts usually obstruct progress. They don't seem to understand that you never get all the "facts". Speed is often just as important or even more important than pinpoint accuracy.

At 6/11/2010 6:39 AM, Anonymous geoih said...

So, if the study shows that any perceived differences in math abilities between males and females are simply artifacts of the test being used to do the measuring, then there must be some other reason (cultural, institutional, psychological, biological, etc.) why females still do not go into math, science and engineering careers.

There are massive efforts put into combatting cultural and institutional obstacles for females going into these careers, yet there is little if any improvement. I have to wonder how the psychological and biological obstacles might be combatted.

At 6/11/2010 9:27 AM, Blogger Junkyard_hawg1985 said...

The arguement made by the reader has one HUGE flaw. If boys outperform girls because they are risk-takers and guess more, this wouldn't explain the huge discepancy between the number of boys and girls who don't miss a single question on the math portion of the SAT. A perfect score does not come about because of "lucky guessing." I did not miss a single math question on the SAT as a senior and it was not because of guessing. Before seeing my score, I expected an 800 because I knew the answers to the questions. I was just hoping I didn't make a simple mistake along the way.

At 6/11/2010 10:22 AM, Anonymous Nick said...

Eliminate standardized testing across the board. It would save a lot of time and money. Some colleges don't even regard the SAT during the admissions process.

Same goes with public school standardized testing - eliminate it. It's huge burden on the taxpayers that encourages "teaching to the test" and not on the actual curriculum.

At 6/11/2010 11:54 AM, Blogger Colin said...

Flawed argument. Its the same as saying if there was no shot clock, no game clock and no pressure to win a basketball game against boys, girls would win.

At 6/11/2010 10:26 PM, Anonymous Anonymous said...

Hi folks! I am the author of the comment featured in the original post. Here's a link where you can find further information:

Gender Bias in College Admissions Tests. (2009). Retrieved from

Also, I wrote some more information about the different ways that the SAT is rigged against girls such as more "masculine" themed questions that boys do better at, the guessing penalty that favors' boys' preference for risk-taking, and the time limit that lowers' girls scores but barely affects boys' scores:

At 6/12/2010 2:43 AM, Anonymous Anonymous said...

Also, I wrote some more information about the different ways that the SAT is rigged against girls such as more "masculine" themed questions that boys do better at

I've seen little on this subject as astonishingly stupid as this. One of the important skills the SAT tests is the ability to distinguish the essential points of a question from the irrelevant details. If a girl can't dissect a word problem because of its "masculine theme", it's the girl's understanding of word problems that's in error, not the problem.

the time limit that lowers' girls scores but barely affects boys' scores

We should care why? Not everything that girls don't do as well at constitute sexism. Time limits don't seem to affect their test grades in school. You know, with its standard hour or so time limit for tests.

Test time limits have a purpose other than easing administration of the test. They are a microcosm of the adult world, where your boss tells you to finish what you're doing by tomorrow, not whenever you feel like it.

At 6/12/2010 8:01 AM, Anonymous Anonymous said...

I wonder how the average SAT math score of the daughters of ardent feminists compares to boys. Just wonderin', you know.


Post a Comment

<< Home