Gender bias: Complexity and Gender bias: Complexity and subtlety in patterns of subtlety in patterns of
stereotyping and prejudicestereotyping and prejudice
Monica BiernatMonica BiernatDepartment of PsychologyDepartment of Psychology
University of KansasUniversity of Kansas
% Female BS Recipients v. Faculty % Female BS Recipients v. Faculty (from Nelson, 2007; data from 2000-2002)(from Nelson, 2007; data from 2000-2002)
0
10
20
30
40
50
60
70
80
% w
om
en .
BS grads
Faculty
% Female PhDs v. Asst Professors % Female PhDs v. Asst Professors (from Nelson, 2007; data from 1993-2002)(from Nelson, 2007; data from 1993-2002)
0
10
20
30
40
50
60
70
80
% w
om
en .
PhDs
Asst Profs
% Female faculty within each rank % Female faculty within each rank (from Nelson, 2007; data from 2002)(from Nelson, 2007; data from 2002)
0
10
20
30
40
50
60
70
80
% w
om
en .
Asst
Assoc
Full
Gender salary gap by academic rank Gender salary gap by academic rank (from Ginther, 2007, data from 2001 SDR)(from Ginther, 2007, data from 2001 SDR)
Reasons for these patterns are manyReasons for these patterns are many– Lack of role models/encouragement at all levelsLack of role models/encouragement at all levels– Lesser access to networksLesser access to networks– Childcare responsibilitiesChildcare responsibilities– Dual career issuesDual career issues
But least some gender discrepancies may be But least some gender discrepancies may be due to gender stereotypes and consequent due to gender stereotypes and consequent gender bias on the part of decision makersgender bias on the part of decision makers
Goals of this talkGoals of this talk
Review experimental research that Review experimental research that documents gender stereotyping effects at documents gender stereotyping effects at all stages of information processingall stages of information processingDocument more complex, counterintuitive Document more complex, counterintuitive effects of gender stereotypeseffects of gender stereotypesNote that these effects can emerge Note that these effects can emerge without conscious intent, awareness, or ill without conscious intent, awareness, or ill willwillDiscuss possible solutionsDiscuss possible solutions
A tour of gender stereotyping A tour of gender stereotyping effectseffects
Automatic gender associationsAutomatic gender associations
Encoding biasesEncoding biases
Construal/memory for ambiguous informationConstrual/memory for ambiguous information
JudgmentJudgment
AttributionAttribution
Definitions of meritDefinitions of merit
Shifting evidentiary standardsShifting evidentiary standards
Content of gender stereotypesContent of gender stereotypes
Women = Communal (Warm)Women = Communal (Warm)– helpful, friendly, kind, sympathetic, helpful, friendly, kind, sympathetic,
interpersonally sensitiveinterpersonally sensitive
Men = Agentic (Competent)Men = Agentic (Competent)– aggressive, ambitious, dominant, self-aggressive, ambitious, dominant, self-
confident, self-reliantconfident, self-reliant
(Williams & Best, 1990; Deaux & Kite, 1993)(Williams & Best, 1990; Deaux & Kite, 1993)
1. We automatically associate these traits 1. We automatically associate these traits with gender categories with gender categories ((Rudman & Glick, 2001)Rudman & Glick, 2001)
-0.4
-0.2
-0
0.2
0.4
0.6
0.8
Strength
of gender-ste
reoty
ped
associa
tions
Women Men
Respondent sex
2. Gender stereotypes lead to 2. Gender stereotypes lead to encoding biasencoding bias
Seeing what we expect to see Seeing what we expect to see
Taking longer to see what we don’t expectTaking longer to see what we don’t expect
Imagine these behavior Imagine these behavior descriptions:descriptions:
““Suzanne works relentlessly to solve difficult Suzanne works relentlessly to solve difficult problems”problems”““William works relentlessly to solve difficult William works relentlessly to solve difficult problems”problems”– Agentic trait: DeterminationAgentic trait: Determination
““Jill encourages employees to approach her if a Jill encourages employees to approach her if a problem arises”problem arises”““Russell encourages employees to approach Russell encourages employees to approach him if a problem arises”him if a problem arises”– Communal trait: UnderstandingCommunal trait: Understanding
How quickly is the relevant trait inferred?How quickly is the relevant trait inferred?
Respondents are slower to infer agentic traits from Respondents are slower to infer agentic traits from women’s agentic behaviorwomen’s agentic behavior (Scott & Brown, 2006)(Scott & Brown, 2006)
540
560
580
600
620
640
660
Time to "encode"
trait
Agentic Communal
Types of behavior exhibited
Female
Male
3. Stereotypes affect construal 3. Stereotypes affect construal of informationof information
Vague or ambiguous information is Vague or ambiguous information is construed to be consistent with the construed to be consistent with the stereotypestereotype
Jane/Bill administered medicine to the patient.Jane/Bill administered medicine to the patient.– Jane/Bill, Jane/Bill, the nursethe nurse, administered medicine to the , administered medicine to the
patientpatient– Jane/Bill, Jane/Bill, the doctorthe doctor, administered medicine to the , administered medicine to the
patientpatient
Elizabeth/Bob was not very surprised upon Elizabeth/Bob was not very surprised upon receiving her/his math SAT score.receiving her/his math SAT score.– Elizabeth/Bob was not very surprised upon receiving Elizabeth/Bob was not very surprised upon receiving
her her lowlow math SAT scoremath SAT score– Elizabeth/Bob was not very surprised upon receiving Elizabeth/Bob was not very surprised upon receiving
his his highhigh math SAT scoremath SAT score
(Dunning & Sherman, 1997)(Dunning & Sherman, 1997)
Memory is stereotype consistent:Memory is stereotype consistent:
– Jane Jane the nursethe nurse, administered medicine to the patient, administered medicine to the patient– Bill, Bill, the doctorthe doctor, administered medicine to the patient, administered medicine to the patient
– Elizabeth was not very surprised upon receiving her Elizabeth was not very surprised upon receiving her lowlow math SAT scoremath SAT score
– Bob was not very surprised upon receiving his Bob was not very surprised upon receiving his highhigh math SAT scoremath SAT score
(Dunning & Sherman, 1997)(Dunning & Sherman, 1997)
4. Gender stereotypes affect 4. Gender stereotypes affect judgments of individualsjudgments of individuals
We judge individual men and women We judge individual men and women consistently with group stereotypes consistently with group stereotypes ((assimilationassimilation))
Biernat & Kobrynowicz (1997)Biernat & Kobrynowicz (1997)
Participants review resume of male or Participants review resume of male or female applicant female applicant
Job is masculine (“chief of staff”) or Job is masculine (“chief of staff”) or feminine (“executive secretary”)feminine (“executive secretary”)
Perceived competence/hireability Perceived competence/hireability assessedassessed
Gender that fits the job is judged Gender that fits the job is judged most competentmost competent
1
2
3
4
5
6
7
Com
pet
ence
.
Chief of staff Secretary
Female Male
Evaluation of CVs Evaluation of CVs (Steinpreis, Anders, & Ritzke, 1999)(Steinpreis, Anders, & Ritzke, 1999)
Academic psychologists evaluate CV of Academic psychologists evaluate CV of biopsychologist Karen Miller/Brian Millerbiopsychologist Karen Miller/Brian Miller
0
10
20
30
40
50
60
70
80
% h
ire
.
Karen Brian
4b. Failure to recognize female 4b. Failure to recognize female expertiseexpertise
Experts identified in group decision-Experts identified in group decision-making task based on actual individual making task based on actual individual performanceperformanceGroup members then interact to reach Group members then interact to reach decisiondecision
Thomas-Hunt & Phillips (2004)Thomas-Hunt & Phillips (2004)
Peer judgments of expertisePeer judgments of expertise
3
4
5
6
7
Experts Non-experts
Women
Men
Women “experts” are judged less expert than men, and even less expert than women non-experts!
Groups with male experts outperformed Groups with male experts outperformed those with female expertsthose with female experts
Why? Because their expertise was usedWhy? Because their expertise was used
Expert Gender (0=F, 1=M)
Influence of expert
Team performance
.11 (.26)
.24 .56
5. Double-bind for women5. Double-bind for women
Women expected to be communal and Women expected to be communal and non-agenticnon-agentic
Perception of competence requires Perception of competence requires agencyagency
Women who display agency may be Women who display agency may be criticizedcriticized
5a. Backlash against self-promoting 5a. Backlash against self-promoting women women (Rudman, 1998)(Rudman, 1998)
3
3.5
4
4.5
5
5.5
6
6.5
Hirea
bility
.
Female Target Male Target
Self-effacing
Self-promoting
5b. Evaluations of men and women 5b. Evaluations of men and women who negotiate for higher salary/benefitswho negotiate for higher salary/benefits
(Bowles, Babcock, & Lei, 2007)(Bowles, Babcock, & Lei, 2007)
4
4.5
5
5.5
6
6.5
7
Hireability .
Accepts Negotiates
Female
Male
Use of self-promotion makes Use of self-promotion makes female female employeesemployees less likeable (B= -.43) and less less likeable (B= -.43) and less competent (B= -.22)competent (B= -.22)
But makes But makes male employeesmale employees more likeable more likeable (B= +.16), and more competent (B= +.34)(B= +.16), and more competent (B= +.34)
5c. Assumption that “successful” female 5c. Assumption that “successful” female managers are not communal managers are not communal (Heilman & Okimoto, 2007)(Heilman & Okimoto, 2007)
Rated desirability of a female or male Rated desirability of a female or male “successful manager” as a boss“successful manager” as a boss
3
3.5
4
4.5
5
5.5
6
6.5
Female
Male
We may attribute stereotype-We may attribute stereotype-inconsistent information to temporary inconsistent information to temporary and/or situational causesand/or situational causes
6. Stereotypes affect attributions for 6. Stereotypes affect attributions for performanceperformance
““What is skill for the male is luck for What is skill for the male is luck for the female” the female” (Deaux & Emswiller, 1974)(Deaux & Emswiller, 1974)
Participants asked to explain the successful Participants asked to explain the successful performance of man or woman on performance of man or woman on “masculine” task“masculine” task
Attribution to ability (versus luck)Attribution to ability (versus luck)
7
7.5
8
8.5
9
9.5
10
Luc
k - A
bilit
yAtt
ribu
tion
FemaleMale
More recent attribution researchMore recent attribution research
Heilman & HaynesHeilman & Haynes (2005) (2005)
Ps read about a successful work team Ps read about a successful work team (one male, one female)(one male, one female)
Judgments of influence/competence of Judgments of influence/competence of group membersgroup members
Male given more credit for team Male given more credit for team successsuccess
1
2
3
4
5
6
7
8
Influential Competent Leader
FemaleMale
7. Stereotypes may affect 7. Stereotypes may affect definitions of meritdefinitions of merit
Emphasizing the importance of attributes a Emphasizing the importance of attributes a favored target possessesfavored target possesses
Evaluation of male or female Evaluation of male or female applicant for police chiefapplicant for police chief– QualificationsQualifications
““Street smart” but not formally educatedStreet smart” but not formally educated
Formally educated, but not “street smart”Formally educated, but not “street smart”
Uhlmann & Cohen (2005)Uhlmann & Cohen (2005)
How important is formal education to being How important is formal education to being a police chief?a police chief?
6
6.5
7
7.5
8
8.5
9
Educated Streetwise
Female
Male
Qualification is more important if the male has that qualification
All of these effects indicate subtle (or not All of these effects indicate subtle (or not so subtle) biases that lead to stereotypical so subtle) biases that lead to stereotypical perceptions of women and menperceptions of women and men
But sometimes stereotypes lead to But sometimes stereotypes lead to contrast effectscontrast effects
Female targets are judged Female targets are judged lessless dependent dependent than males than males (Corneille et al., 2000)(Corneille et al., 2000)
Women are judged to be Women are judged to be more agentic more agentic than men than men (Eagly & Steffen, 1984)(Eagly & Steffen, 1984)
These These contrast contrast effects may also be effects may also be driven by stereotypesdriven by stereotypes
8. Stereotypes activate shifting 8. Stereotypes activate shifting standards standards (Biernat, 2003; Biernat & Manis, 1994)(Biernat, 2003; Biernat & Manis, 1994)
Stereotypes serve as standards against Stereotypes serve as standards against which we judge individual group members.which we judge individual group members.
Standards Standards shiftshift when we judge members when we judge members of different groups.of different groups.
Based on stereotypes . . .Based on stereotypes . . .
How tall is tall?How tall is tall?– For men, 6’2”, for women, 5’9”For men, 6’2”, for women, 5’9”
How aggressive is aggressive?How aggressive is aggressive?– Standard is lower for womenStandard is lower for women
How emotional is emotional?How emotional is emotional?– Standard is lower for menStandard is lower for men
ImplicationsImplications
Identical behavior is judged differently Identical behavior is judged differently depending on who does itdepending on who does it
Identical judgments may mask very Identical judgments may mask very different underlying perceptionsdifferent underlying perceptions
Gender and judgments of financial success Gender and judgments of financial success (Biernat et al., 1991)(Biernat et al., 1991)
Ps view 40 photos of M and F targetsPs view 40 photos of M and F targets
Judge financial success in:Judge financial success in:– Common rule units ($ earned per year) orCommon rule units ($ earned per year) or– Subjective units (1-7 Subjective units (1-7 unsuccessful-successfulunsuccessful-successful))
Financial success judgmentsFinancial success judgments
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
Stdi
zed
fina
ncia
l suc
cess
.
Dollars Subjective
Females
Males
Standards may shift in the direction Standards may shift in the direction of stringency or leniencyof stringency or leniency
Standards may be Standards may be lowerlower or or higher higher for for groups stereotyped as deficient on an groups stereotyped as deficient on an attributeattribute
Stephen Carter (1993) Stephen Carter (1993) Reflections of an Reflections of an affirmative action babyaffirmative action baby
““Our parents’ advice was true: We really do Our parents’ advice was true: We really do have to work twice as hard to be considered have to work twice as hard to be considered half as good (as Whites)” (p. 58)half as good (as Whites)” (p. 58)
““(I am) somewhat skeptical that Black (I am) somewhat skeptical that Black people of intellectual talent have a harder people of intellectual talent have a harder time than others in proving their worth. My time than others in proving their worth. My own experience suggests quite the contrary, own experience suggests quite the contrary, that like a flower blooming in winter, intellect that like a flower blooming in winter, intellect is more readily noticed where it is not is more readily noticed where it is not expected to be found” (p. 54)expected to be found” (p. 54)
Types of standardsTypes of standards
Minimum: Standard to suspect that Minimum: Standard to suspect that someone has an attribute; expectationsomeone has an attribute; expectation
Confirmatory: Standard to confirm that Confirmatory: Standard to confirm that someone has attributesomeone has attribute
Minimum standards are lower but Minimum standards are lower but confirmatory standards are higher for confirmatory standards are higher for groups stereotyped as deficient on groups stereotyped as deficient on attributeattribute
Work-related competence standards: Work-related competence standards: Women stereotyped as deficient Women stereotyped as deficient
(Biernat & Kobrynowicz, 1997)(Biernat & Kobrynowicz, 1997)
2.58
3.12
3.66
2.91
1
1.5
2
2.5
3
3.5
4
Ski
ll e
xam
ple
s to
mee
t st
d .
Minimum Confirmatory
Standard
Katherine
Kenneth
Gender and short-listing/hiring Gender and short-listing/hiring (Biernat & Fuegen, 2001)(Biernat & Fuegen, 2001)
82
66
39
46
20
30
40
50
60
70
80
90
Per
cent
yes
.
Short list Hire
Decision
KatherineKenneth
9. Differential standards may lead 9. Differential standards may lead to different behavioral rewardsto different behavioral rewards
Men given consequential rewards; women Men given consequential rewards; women given praise for “good” performancegiven praise for “good” performance
Co-ed softball and standards Co-ed softball and standards (Biernat & Vescio, 2002)(Biernat & Vescio, 2002)
Role-playing managers favor men in Role-playing managers favor men in assigning positionsassigning positions– Team selections:Team selections: 52% men52% men– Benching decisions:Benching decisions: 59% women59% women– Infield positions:Infield positions: 58% men58% men– Top of batting order:Top of batting order: 63% men63% men
But cheer more in response to a woman’s But cheer more in response to a woman’s getting on basegetting on base
Praise but no raise in a work setting Praise but no raise in a work setting (Vescio et al., 2005)(Vescio et al., 2005)
Male team leaders assigned women to Male team leaders assigned women to fewer valued positions than men, but fewer valued positions than men, but praised them more for their work productpraised them more for their work product
Women treated this way reported more Women treated this way reported more anger and worse subsequent performanceanger and worse subsequent performance
Summarizing effects of Summarizing effects of stereotypesstereotypes
Quick associationsQuick associationsEncoding biasesEncoding biasesConstrual/memory for ambiguous informationConstrual/memory for ambiguous informationJudgment – assimilation to stereotypesJudgment – assimilation to stereotypesThe double bindThe double bindAttributionAttributionDefinitions of meritDefinitions of meritContrast effect in evaluationContrast effect in evaluationEvidentiary standards – stringency and leniencyEvidentiary standards – stringency and leniency
What can we do about gender What can we do about gender bias?bias?
Awareness and conscious self-correction Awareness and conscious self-correction can helpcan help– MonitoringMonitoring– TrainingTraining
Acknowledgement of subtle biasAcknowledgement of subtle bias
Changing associations to gender and to job Changing associations to gender and to job categoriescategories
Suppressing/controlling/correctingSuppressing/controlling/correcting
Changing normsChanging norms
University of Wisconsin search University of Wisconsin search committee training committee training (Pribenow, Maidl, & Winchell, 2005)(Pribenow, Maidl, & Winchell, 2005)
Workshops on good search practices + effects Workshops on good search practices + effects of nonconscious gender/race biasof nonconscious gender/race bias
05
1015202530354045
% w
om
en a
sst pro
f hires .
Participating depts Non-participatingdepts
1999-2002
2003-2005
Accountability for decisionsAccountability for decisions– Search processesSearch processes– Tenure decisionsTenure decisions
Gender-blind review processGender-blind review process– Journal articles, grant proposalsJournal articles, grant proposals
Family-friendly policiesFamily-friendly policies– In definition of tenure clock and career In definition of tenure clock and career
trajectorytrajectory