f:\doc\web\2000\11\stereo.txt
Date: Thu, 02 Nov 2000 12:44:37 -0500 
From: "DeLauro, Linda" 
Subject: Studies Concerning Stereotype Threat 
To: "'jheckman@midway.uchicago.edu'" 
Cc: "Stricker, Larry" 
X-Mailer: Internet Mail Service (5.5.2650.21) 


Steele, his co-workers, and others have done a great many laboratory
studies on different aspects of stereotype threat with different
populations. I have done only four studies (one not yet written up),
all with ETS tests and ETS test takers.


Two of my studies, stemming from and extending or extrapolating a
Steele and Aronson study, concerned whether asking about ethnicity
and gender affected performance on two operational tests in a field
experiment: the Advanced Placement test in calculus, taken by high
school students, and the Computerized Placement Test Battery (a
measure of basic skills), taken by community college students. We
found no differences for minority test takers or women that were both
statistically and practically significant (accounting for at least 1%
of the variance) in our large samples.


A third study, a close replication of one by Spencer, Steele, and
Quinn, concerned whether making a test easier reduced the effects of
stereotype threat. This was a laboratory experiment with the GRE
General Test, using college seniors bound for graduate school or
first-year graduate students. We found no statistically and
practically significant differences for Black or female test takers
in our large sample.


The fourth study, analyzed but not written up yet, concerned a
general assumption underlying stereotype threat theory: minority and
female test takers think that the majority group believes they have
poor intellectual ability (poor quantitative ability in the case of
women). This was a survey, with attitude questions appended to an
operational Graduate Management Admission test for several months.
Two questions asked: When other people evaluate your verbal
(quantitative) ability, do you think their estimates are: Much too
high ... much too low? We found no statistically and practically
significant differences among the ethnic groups or the sexes in our
enormous sample (over 120,000).


Larry Stricker


Linda J. DeLauro
ETS-GRE Collaborative Validity Study
Project Manager
Mail Stop 09R
Princeton, NJ 08541
Email: ldelauro@ets.org
Tel: 609-734-1806
FAX: 609-734-1755




James J Heckman
Dept of Economics
University of Chicago
1126 E 59th Street Chicago ILL 60637 USA
fax 773-702-8490 phone 773-702-0634 email jjh@uchicago.edu