Quote:
|
Originally Posted by The LD
I'm not sure that you don't have the cart in front of the horse here, to be honest. Maybe you have the studies or information that can resolve this, but the assumption that you make here is that grading in math courses is the "true" measure of mathematical aptitude, and that the SAT test is somehow skewed. The SAT is a series of math questions to be performed under pressure, nothing more. College courses, particularly survey-level courses, often include factors like points for turning homework in on time, bonus credit for additional work, etc. Why should we believe that this is the accurate measure?
|
No, I'm assuming that there isn’t any "true" measure of aptitude as how you measure something affects what it is you are measuring. For example, there is research by Sharps and his colleagues that show that gender differences on certain mathematical tasks depend on how the task is described to participants. It seems like common sense, but people tend to do well on tasks that are familiar to them. In his study, Sharps found that men performed much better than women when the instructions emphasized the usefulness of these mathematical abilities in stereotypically masculine professions (e.g., engineering). When the instructions emphasized how these abilities were useful in stereotypically feminine occupations, the gender difference in the math scores disappeared. So given that men and women may tackle the same problem in different ways and depending on how it is asked, it makes sense that the SAT should ask questions that assess mathematical problems from various angles. However, the SAT was created by - and tested on - white men. And while I don’t think this was a conscious bias, the end result is that people who are not white men are scoring significantly lower on these standardized tests.
Further, the reason why I brought this up in the first place is that opponents of Affirmative Action argue that admissions to colleges should be based on these standardized tests – tests which routinely show differences in scores among various groups. The SATs stated purpose is that it is useful for predicting first-year college grades. The study I cited shows that it is not useful for predicting women’s first year grades and there are other studies showing it is not useful for predicting ethnic minority grades either. As such, the SAT and other "standardized" tests should not be used as a main criteria for admissions until they can be proven to measure people's abilities equally.
Quote:
| Moreover, why does this disparity is SAT scores exist in the first place? I'm going to go ahead and say I find it ridiculously implausible that the math questions are designed to create an advantage for males. So, something else is at play. Are women affected by the stress of timing? Do men thrive under such pressure? Something is at play here, and simply saying "it doesn't reflect their grades that well" (which is a conclusion I also have a problem with, as simple linear programming could easily rectify the situation) doesn't get rid of the aberrant results. |
Well, there are tons of theories about this - enough to fill years’ worth of course work. But a summary of possible gender bias can be found here:
http://www.fairtest.org/facts/genderbias.htm
To bring the discussion back to racial disparities, some researchers believe stereotype threat can explain these differences. Claude Steele first showed this effect in his studies on Black students. According to stereotype theory, if you belong to a group that is hampered by negative stereotypes – and you are reminded about your membership in that group – your performance suffers. One stereotype about Black people is that they are academically incompetent particularly in comparison to Whites. In his study, Steele told one group of students (both Black and White) that the test was designed to test their academic ability. He told another group of students (both Black and White) nothing about what it was supposed to measure. The White students performed almost equally in the two conditions of the experiment. Blacks, in contrast, performed far worse than they otherwise would have when they were told their intelligence was being measured. Further experiments manipulated how stereotypes were primed. For example, sometimes students were asked to fill out a demographic sheet that specifically asked about race prior to taking an exam. Other experiments manipulated whether the test was administered by a Black or White staff person. When race is made salient in a variety of different ways Black students performed worse. To show that this effect occurs for a variety of different stereotypes, these studies have been tested on women and various ethnic groups. All the studies show that when a person is reminded that they are a member of a particularly devalued group, their performance suffers.
One study looked at what happens when people have competing stereotypes. The stereotype for women is that they are bad at math. The stereotype for Asians is that they are good at math. So what happens when we look at mathematic test scores of Asian American women? Shih and her colleges experimentally manipulated whether Asian American women were reminded of their Asian status or their gender right before taking a math test. There was also a control group in which the students did not have to fill out a questionnaire about their race or gender. Asian women who were reminded of their ethnicity performed well; however women who were reminded of their gender performed poorly. The control group performed on par with the average for other students.
Typically women who go to all girls’ schools or Blacks who go to historically Black universities tend to do very well academically. Now I don’t want to advocate segregation here. I just want to point out that biologically and genetically women and people of color aren’t inherently inferior in mathematic or other academic abilities. There’s something else going on and while there are many theories about what that something else may be, we don’t have conclusive evidence one way or the other.
One other thing I want to point out are the socialization theories. Studies of very young children show that boys and girls of all races perform equally on a number of academic and social tasks. But once kids enter school we start seeing differences. There is lots of evidence that people respond to boys and girls differently, as well as treat people from various ethnic groups differentially, and this biased treatment is not conscious. For example, studies show that girls and people of color are largely ignored in the classroom. Teachers typically pay more attention to boys than girls; boys receive more positive and more specific feedback; and boys are more likely to be recognized, called on in class, and included in class discussions. All this combines to encourage boys to think of themselves as academically competent, particularly in stereotypically masculine areas (i.e., math and science).
In one study Basow (2004) found that in elementary school, Black girls spoke up in the classroom and were quite engaged in learning. But their assertiveness was discouraged by their teachers - by both verbal and nonverbal cues. Basow tracked these girls to the fourth grade and found them to be significantly more passive and quiet. In addition, none of these girls were asked to take on more academic responsibilities, such as tutoring or showing a new student how to prepare an assignment. These Black girls were socialized to believe that school is not for them.
I feel like I'm just spouting studies now, but the questions you bring up LD are questions that researchers have been tackling for decades. It's hard to just sum that research up in a tidy little statement. But back to my original point, schools and institutions have the ability to reduce stereotypes and bias by recognizing how gender and race can influence attitudes and behaviors of all individuals, as well as the judgments people make about our own and others’ abilities. Until admission offices can separate out individual ability from gender, race, class, and a whole host of other social issues, those issues need to be taken into account. Again, I'm NOT advocating that grades and test scores don't mean anything. Clearly a woman or person of color who scores in the bottom percentile of any standardized test is not as qualified as someone who scores in the top percentile. But if schools use an absolute cut off score such as a 630 on the math SAT and women on average are scoring 600 (compared to 633 for men), you can see how that makes a difference in who gets accepted or not.