I’ve never been very impressed with Claude Steele’s theory of “stereotype threat (as indicated here), and a new study by Prof. Geoffrey Cohen at the University of Colorado and a colleague at Yale that is getting some attention hasn’t changed my mind.
The study involved two experiments in which seventh graders from middle-class and lower middle-class families were asked to choose one or two values that were important to them and then to write a paragraph describing why they cherished the values. A control group was asked to write about values that others might hold. A total of 243 students participated, divided in roughly equal numbers by race.
“This exercise, called a self-affirmation, allows a student to reaffirm that he or she is a good person,” Cohen said. “That helps fight the stress arising from the fear that the negative stereotype could be used against you.”
The study found the average performance gap in the course between African-American students and their white peers at a suburban middle school in the northeastern United States was three-quarters of a grade point on a four-point scale for those in the control group.
The African-American students who completed the in-class assignment improved their end-of-term grades by three-tenths of a grade point, closing the gap by 40 percent, according to Cohen. The assignment had no impact on white students’ grades.
Actually, this description of the study, from a University of Colorado News Release, is not accurate, or at least it differs in a material way from the description contained in a long article today in the Chronicle of Higher Education. According to that article, only the grades of the black students who had written about values that were important to them improved; the grades of the other black students, and the white students, were unchanged:
The students completed the assignment in class at the beginning of the fall term, and the researchers followed the progress of those students. At the end of the year, African-American students who had written about their most important values had better grades than did African-American students in the control group. The difference was about one-third of a grade point on a four-point scale, where an A is a 4 and a D is a 1.
White students who wrote about their most important values did not show any significant difference than white students in the control group.
The Chronicle also gave the following examples:
From African-American students who were asked to write about why certain values were important to them.
• “My friends and family are most important to me when I have a difficult situation that needs to be talked about. My friends give me companionship and courage. My family gives me love and understanding.” (female)
• “Well being a great athlete and hitting the book are really the most important things in my life. I’m a great athlete when it comes to sports like basketball and football but when it comes to school I try and try to work as hard as I can to go to college and to make my family proud.” (male)
From African-American students in the control group, who were asked to describe why values that were least important to them would matter to another person.
• “Athletic abilities may be important to someone who comes from an athletic family. They probably feel that everyone wants them to live up to the capabilities of your family member(s). It may be important to someone else because they are trying to live up to your dream of becoming a football player, basketball player or whatever. This is not important to me because I want to be a pediatrician or lawyer.” (female)
• “This value [being good at art] would be important to someone else because they might be good at that. They might best at it or the might be happy when they do it.” (male)
I wouldn’t argue that these findings are incorrect. Maybe it’s true that “self-affirmation” works, that the grades of a 7th grade boy who writes that he’s a great athlete and tries hard in school will improve somewhat while those of a 7th grade girl who writes that athletic ability is not important to her because she wants to be a pediatrician or a lawyer will not.
But if so, if the boy’s lower initial grades were the result of “stereotype threat,” and if the effects of stereotype threat can be reduced by 40% by writing a short comment about an important value, then that threat strikes me as pretty weak in the first place.
UPDATE [6 Sept.]
Linda Seebach pointed me to an excellent critique of this study. After some rigorous, informed statistical analysis and criticism, the critique concludes that the confidence level of the findings is “close enough to bupkis for government work,” that the selection of students and classes “smells fishy,” and the problem of finding some solutions to the black–white achievement gap is difficult but that “we’ll only find one by acknowledging the truth and thinking up thick-skinned solutions rather than writing letters to Santa Claus.”
John,
It’s the same basic way Patricia Gurin “proved” diversity has educational benefits.
She studied her own students, whom she fed a curriculm, and then after receiving that curriculm, she compared their responses to before they started on the curriculm. She found – lo and behold – that the students performed measurably “better” as she defined the word “better” (she actually defined “democracy outcomes” as including support for environmental policies, among other things).
She admits that isolating numerical diversity alone does nothing – it must be combined with indoctrination (of course, if we can’t isolate it and measure it, then it ain’t science, as they say).
It’s an interesting study. I’m not sure the right conclusions are being reached. I think the results are interesting enough to warrant some open-minded speculation. For instance, maybe the 40% improvement was simply a matter of having the kids focus on and articulate their goals.
I wrote here about my definition of success, which is acheivement of goals; and how it is important to review one’s goals periodically to make sure one is on track for meeting them. I worked this definition out in the process of expressing some things to my then-high-school-age daughter. Because I’ve always talked to her about stuff like this. Maybe some parents don’t talk to their kids about defining their values, and if some caring teacher does and the kids’ grades go up as a result, that’s something to take notice of and build on.
The study probably hasn’t found anything, but even if they have, it might be just that black 7th graders are many times more likely to go through extreme flippings from severely anomic conditions, to one in which their grades can bounce up enough to get some A’s for effort.
if you gave kids ice cream before a test, no doubt some would perform better that if they had not had the ice cream. how can schools or testers or anyone plan or account for every possible thing that may affect a child’s performance.
I know some one who has this problem. I have know her all my life and I think that all of this could change a childs life but that is just my opinion.