Overcoming Hidden Bias
RECOGNIZING THE BEST AND THE
BRIGHTEST
Who are we? Why are we here?
Prof. Thomas DeFrantz
Professor of Music and Theater Arts
Director, Women’s and Gender Studies
Prof. Sally Haslanger
Professor of Linguistics and Philosophy
Outline
•
•
•
•
•
•
Why might we care about diversity?
Some data
Schemas and implicit bias
Stereotype threat and solo status
Feedback loops
Discussion
An annotated bibliography and copies of research we are relying on can be
found on Stellar at:
http://stellar.mit.edu/S/project/equityissues/index.html
MIT’s mission statement
The Institute is committed to generating, disseminating,
and preserving knowledge, and to working with others to
bring this knowledge to bear on the world's great
challenges. MIT is dedicated to providing its students with
an education that combines rigorous academic study and
the excitement of discovery with the support and
intellectual stimulation of a diverse campus community. We
seek to develop in each member of the MIT community the
ability and passion to work wisely, creatively, and effectively
for the betterment of humankind.
Why do we care?
• We want the highest quality faculty we can get to
respond to the conditions of the world today.
o For excellence in research, competitiveness in our disciplines.
o To respond to the growing diversity of our student body.
• Fairness is crucial in a genuine meritocracy.
• Women and minorities are underrepresented in high
status occupations relative to their quality.
o The composition of the candidate pool only accounts for part of
the problem.
o Research shows that we all – regardless of gender or race –treat
people differently based on their perceived social group.
• All else being equal, diversity is valuable.
5
School of Science Tenure Data
Numbers in EAPS
Schemas and Discrimination
• Explicit Discrimination: conscious actions directed
against members of a group.
• Schemas: Non-conscious expectations or
stereotypes associated with members of a group
that guide perceptions and behaviors.
• Action based on schemas is pervasive and
inevitable. But schemas can be distorting and
result in poor judgment.
Valian (1998) Why So Slow? The Advancement of Women. Cambridge: MIT Press, p. 280
13
Schemas are widely shared
• Research shows that we all – regardless of gender or
race – perceive and treat people based on schemas
associated with their race/gender/social group.
• Both men and women hold them about gender
• Both whites and people of color hold them about race
• People are typically not aware of them, but with effort
can become aware of them and change them.
• Implicit association test:
https://implicit.harvard.edu/implicit/
Fiske (2002). Current Directions in Psychological Science, 11, 123-128.
14
Schemas play a significant role
when there is:
•
•
•
•
Ambiguity (including lack of information)
Stress from competing tasks
Time pressure
Under-representation of the group in question
(when the group does not reach critical mass)
Evaluation of Fellowship Applications
Average rating of applicants as a function of
their scientific productivity
3
2.75
Score
“…the success rate of female scientists
applying for postdoctoral fellowships at
the [Swedish Medical Research Council]
during the 1990s has been less than half
that of male applicants.”
2.5
2.25
2
0-19
Results of study: Women
applying for a post- doctoral
fellowship had to be 2.5 times
more productive to receive the
same reviewer rating as the
average male applicant.
20-39
40-59
60-99
>99
Total impact
males
females
Similar findings:
• USA/GAO report on Peer Review in Federal
Agency Grant Selection (1994)
• European Molecular Biology Organization
Reports (2001)
• NIH Pioneer Awards: Journal of Women’s
Health (2005) & Nature (August 2006)
Wenneras & Wold (1997) Nature, 387, 341.
18
Gender Schemas in Recommendations
for Successful Medical School Faculty Applicants
Letters for men:
• Longer
• More references to CV, Publications,
Patients, Colleagues
Letters for women:
• Shorter
• More references to personal life
• More “doubt raisers,” including hedges, faint
praise, and irrelevancies (e.g., “It’s amazing how
much she’s accomplished.” “It appears her health is
stable.” “She is close to my wife.”)
Trix & Psenka (2003) Discourse & Society, 14(2): 191-220, 2003.
More evidence of evaluation bias
• Race stereotypes often lead to different
standards of assessment.
• Women and minorities are more easily judged
competent,
• But standards for excellence are set higher than
for men and whites.
Biernat & Kobrynowicz, 1997
Stereotype Threat
• Stereotype threat occurs when your group is
stereotyped as performing poorly in a domain
and your performance may appear to confirm
the negative stereotype.
– Performance decreases on computational and recall
tests.
– Conscious awareness of the threat is not necessary
for the effects.
• Stereotype threat is situational: performance
decreases only in settings where the stereotype
is activated.
Solo Status
• Solo status occurs when one is the only
member of one’s social group in a setting.
• Solo status increases the risk of stereotype
threat; public settings also exacerbate the
effects.
• Explanation of stereotype threat is contested,
but addressing solo status can reduce
stereotype threat.
Accumulation of advantage &
disadvantage
• Like interest on capital, advantages accrue.
• Like interest on debt, disadvantages accrue.
• Small differences in treatment can
accumulate to cause major consequences in
salary, promotion, prestige.
Accumulation of disadvantage
feedback loop
LOWERED CAREER SUCCESS
RATE
Affects educational opportunities, publication rate,
funding, climate of support, etc.
“Confirms”
schema
Performance is underestimated
Evaluation Bias
Lack of
critical
mass
Brainstorming
• How do we account for the numbers in your
department?
• What is your explanation?
Brainstorming
• What actions, strategies, solutions do you
have to offer?
Accumulation of disadvantage
feedback loop
Improved
LOWERED CAREER SUCCESS
RATE
Affects educational opportunities, publication rate,
funding, climate of support, etc.
Disconfirms
“Confirms”
schema
Creates
Valued appropriately
Performance is underestimated
Evaluation Bias
Lack of
critical
mass
MIT Diversity Congress
Prof. Michael Summers, UMBC