What's In a Name...

Hey all!  I know summer's a good time for relaxing and all, but lots of studying for other tests (GRE, MCAT, and LSAT, anyone?) has been going on.  That reminds me that I should start studying for the GRE myself.  But before that, I'm also reminded of (another) psych experiment...

Most people know the stereotypes associated with math performance.  According to popular belief, Asians are good at math, but women are bad at it...  But, how well do Asian girls do?  This was one of the questions that Shih, Pittinsky, and Ambady (1999) raised (http://pss.sagepub.com/content/10/1/80.short).

You see, other experiments have shown that stereotype activation has an effect on performance.  Studies have shown that if people are made aware of the stereotypes associated with their characteristics, they'll perform in ways that fit with those stereotypes.  For example, when White and Black students took a test that was "diagnostic of their abilities," Black students scored lower than White students, which is consistent with the stereotype.  Yet, if the test was passed as non-diagnostic (the researchers said that they wanted to see how people solved tough verbal problems), there was no significant difference in the scores of White and Black students.  This is huge because it shows that just even being aware of certain stereotypes can affect one's performance...

So, back to the original study.  All the participants were female Asian college students.  The three conditions were female-identity-salient condition, Asian-identity-salient condition, and no-identity-salient condition (the control).  In each condition, participants had to fill out a survey asking about residential life at their university.  What changed in each condition is that the survey's questions differed for each.  In other words, for the female-identity-salient condition, participants had to indicate their gender and answer questions related to it.  The same idea essentially applies to the Asian-identity-salient condition.  As for the no-identity-salient, participants filled out questions that were about residential life but specifically avoided ethnic or gender identity.  After the survey, all participants took a math test.

For ease of explanation, let's just say that the participants in the no-identity-salient condition scored what others score on average.  In other words, the control condition is like the anchor of which the scores for the two other conditions are judged/based off of.  Some of you may have guessed already, but participants in the gender-salient-identity condition, on average, scored significantly lower than those in the no-identity-salient condition.  In other words, Asian females made aware of their gender tended to score lower than those who did not have their gender or ethnicity mentally reinforced.  Also, participants in the Asian-identity-salient condition, on average, scored significantly higher than those in the no-identity-salient condition.  In other words, Asian females made aware of their ethnicity tender to score higher than those who did not have their gender or ethnicity mentally reinforced.

This is not only quite amazing but disturbing as well.  Sex and ethnicity are two factors that are very innate and are (arguably) difficult to change.  Yet, such concrete factors can have a significant effect upon performance, as witnessed by the previous experiment.  Is there any hope?  Or are those that fit stereotypes doomed to fulfill them..?

Actually, the experiment already proposes a simple solution:  don't make people aware.  It may sound overly simple, but the control seemed to be balanced between the two other conditions, suggesting that it may be the most balanced.  Thus, if people didn't have their gender or ethnicity reinforced, tests may be a more accurate measure of the test taker's innate ability.

Of course, you may think that reinforcement like the surveys found in this experiment ever happen on tests.  So, it shouldn't be a problem in real world testing.  Oh, if it was only the case...  You see, apparently, my Social Psych Laboratory teacher Dr. Goff was telling us that even writing one's ethnicity and/or gender has enough of an effect.  In other words, simply writing "Black" at the top of a test has been linked to Black students scoring lower on tests.  In fact, Goff was telling us that this was really big for the GREs, and the makers of the GRE were/are being sued for refusing to add questions about ethnicity to the end of the test rather than the beginning.

So, there you go...  If something as simple such as writing one's ethinicity or gender can cause a significant difference in test scores, it makes me wonder how accurate the statistics reported that have measures of intelligence for each ethnicity truly is.  Of course, people could say that it could be helpful for certain groups like Asians, for if Asian identity was reinforced, Asians may score higher on average than if they had no identity reinforcement (just like the experiment).  Yet, is it really ethical for a group of people to have an advantage over others just because of such a concrete factors as ethnicity?  And are there more positive or negative stereotypes related to race in general..?  Just some stuff to think about, I guess...  Anyways, keep on enjoying summer, everyone!