Скачать презентацию Best Practices in Faculty Searches Sabrina Burmeister Associate Скачать презентацию Best Practices in Faculty Searches Sabrina Burmeister Associate

bb43db155c887e268326bc31068b8c2b.ppt

  • Количество слайдов: 27

Best Practices in Faculty Searches Sabrina Burmeister Associate Chair for Diversity, Diversity Liaison, Equal Best Practices in Faculty Searches Sabrina Burmeister Associate Chair for Diversity, Diversity Liaison, Equal Opportunity Officer Department of Biology

Best Practices in Faculty Searches Acknowledgements: Sabrina Burmeister (Biology), Kia Caldwell (Director Faculty Diversity Best Practices in Faculty Searches Acknowledgements: Sabrina Burmeister (Biology), Kia Caldwell (Director Faculty Diversity Initiatives), Adrienne Erickcek (Physics & Astronomy), Jeff Johnson (Chemistry), Sheila Kannappan (Physics & Astronomy), and Laurie Mc. Neil (Physics & Astronomy), Keith Payne (Psychology) Abigail Stewart (U. of Michigan ADVANCE program)

Best Practices in Faculty Searches We all have the same goal: A fair process Best Practices in Faculty Searches We all have the same goal: A fair process that efficiently and objectively identifies outstanding candidates.

Why the familiar process falls short We are not computers: When comparing multiple choices Why the familiar process falls short We are not computers: When comparing multiple choices on multiple dimensions simultaneously, we depend on cognitive shortcuts that lead to biased outcomes. Humans are not “rational” decision makers.

Behavioral Economics “Irrational” decision making and cognitive shortcuts Example: Decoy effect House 1 House Behavioral Economics “Irrational” decision making and cognitive shortcuts Example: Decoy effect House 1 House 2 House 3 Colonial Contemporary Colonial $200, 000 needs new roof < $2000 Yes No No

Behavioral Economics “Irrational” decision making and cognitive shortcuts. Example: Decoy effect Many cognitive shortcuts Behavioral Economics “Irrational” decision making and cognitive shortcuts. Example: Decoy effect Many cognitive shortcuts are evolutionarily inherited. Example: Frogs show decoy effect Implicit biases are social constructs (i. e. learned). Examples: Girls are bad at math, leaders are men, fat people are lazy, black people are good athletes.

Implicit (Unconscious) Biases Implicit biases are one of the most robust and replicable phenomena Implicit (Unconscious) Biases Implicit biases are one of the most robust and replicable phenomena in the field of psychology. Implicit bias is not a code word for racist/misogynist. They are a product of our culture: Over time, stereotypes become automatic associations. They create positive feedback loops (e. g. , girls are bad at math: teachers give better math grades to boys; girls are anxious about math, they score lower on standardized tests).

Implicit Biases • Held by everyone, including non-majority groups. (implicit. harvard. edu) e. g. Implicit Biases • Held by everyone, including non-majority groups. (implicit. harvard. edu) e. g. female faculty in Moss-Racusin et al. (2012). • MORE likely among those who consider themselves objective. Not correlated with education. Monin & Miller (2001); Uhlmann & Cohen (2007) • Applied more under circumstances of: • Ambiguity (including missing information) • Stress from competing tasks • Time pressure • Lack of critical mass in applicant pool Fiske (2002). Current Directions in Psychological Science, 11, 123 -128 8

Implicit biases lead to biased outcomes White job candidates 50% more likely to be Implicit biases lead to biased outcomes White job candidates 50% more likely to be interviewed than black job candidates in spite of identical CVs. (Bertrand 2004) White males are more likely to receive positive response to email from faculty compared to females or minorities in spite of identical email content. (Milkman et al. 2012 Psychological Science When evaluating identical undergrad candidates for lab manager position, faculty ranked “John” more hireable, competent, and worth mentoring than “Jennifer. ” (Moss-Racusin et al. (2012). PNAS) When evaluating identical applications, male and female professors preferred 2: 1 to hire “Brian” over “Karen” as an assistant professor. (Steinpreis, Anders, & Ritzke (1999) Sex Roles, 41, 509) 23: 710) 9

Why do we care? A biased process means that we are overlooking talent. We Why do we care? A biased process means that we are overlooking talent. We value diversity – A diverse faculty is better positioned to educate a diverse student body. Diverse groups of people are more innovative and creative when solving problems. Advantages of diverse groups at the small scale have large scale consequences: Diverse companies are more successful financially.

An evidence-based search process 1. Search Committee 2. Job Ad 3. Design rubric(s) 4. An evidence-based search process 1. Search Committee 2. Job Ad 3. Design rubric(s) 4. Recruit actively. 5. Apply a rubric at every stage. 11

1. Search committee Why start with the search committee? The job ad should accurately 1. Search committee Why start with the search committee? The job ad should accurately reflect the search image. The search image should be based on individual qualifications and not on a global ideal. 12

1. Search committee • • Diversify the search committee when possible. Minorities are not 1. Search committee • • Diversify the search committee when possible. Minorities are not less biased, but their presence matters. Study of Racial Diversity in Jury Deliberations: Compared with all-white juries, diverse juries deliberating about an African American defendant: –Spent more time discussing the case –Mentioned more facts –Made fewer inaccurate statements –Left fewer inaccurate statements uncorrected –Discussed more race-related issues Sommers (2006) Journal of Personality and Social Psychology, 90 (4), 597 -612. 13

1. Search committee • • Diversify the search committee when possible. Include a search 1. Search committee • • Diversify the search committee when possible. Include a search advocate. From Oregon State University: Works with the search committee on job ad before the position is posted. Provides research-based information about implicit cognitive and structural biases that affect search and selection processes. Suggests strategies to help mitigate the effects of those biases. The Search Advocate works collaboratively within the group, and on occasion may offer a perspective that helps the committee test its thinking. - 14

1. Search committee • • • Diversify the search committee when possible. Include a 1. Search committee • • • Diversify the search committee when possible. Include a search advocate. Provide implicit bias training to committee. **Implicit bias training does not eliminate their effects. Knowledge of implicit bias enables us to embrace practices that reduce their effects. 15

2. Job Ad • Words that match male stereotypes (e. g. , leader, competitive, 2. Job Ad • Words that match male stereotypes (e. g. , leader, competitive, rigorous) discourage female applicants. Words that match female stereotypes (e. g. , collaborative) encourage female applicants. Gaucher, et al. 2011 J. Personality and Social Psychology 101: 109 -128. • Communicating that we value diversity increases the number of minority applicants: e. g. , “We seek applicants who engender a climate that values diversity in all its forms. ” Smith, 2004 Interrupting the Usual • Require statement of Contributions to Diversity as part of application process. http: //facultyexcellence. ucsd. edu/c 2 d/index. html 16

2. Job Ad Contributions to Diversity statement: “The Contributions to Diversity Statement should describe 2. Job Ad Contributions to Diversity statement: “The Contributions to Diversity Statement should describe your past experience, activities and future plans to advance diversity, equity and inclusion, in alignment with UC San Diego’s mission to reflect the diversity of California and to meet the educational needs and interests of its diverse population. ” http: //facultyexcellence. ucsd. edu/c 2 d/index. html 17

An evidence-based search process 1. Search Committee 2. Job Ad 3. Design rubric(s) 4. An evidence-based search process 1. Search Committee 2. Job Ad 3. Design rubric(s) 4. Recruit actively. 5. Apply a rubric at every stage. 18

3. Rubric Design UNC policy (http: //www. unc. edu/depts/eooada/sct/index. htm): 3. Rubric Design UNC policy (http: //www. unc. edu/depts/eooada/sct/index. htm): "The selection criteria must be carefully defined, directly related to the requirements of the position, and clearly understood and accepted by members of the search committee. The ability of the candidate to add intellectual diversity and cultural richness to the department should be included among the selection criteria. ” (pg. 26) 19

3. Rubric Design • Avoid global judgments • Abandon the myth of the “best 3. Rubric Design • Avoid global judgments • Abandon the myth of the “best candidate” – this concept increases the probability of a biased outcome. • Actuarial (Fact-based) Judgment always outperforms Clinical (Professional) Judgment • Use fact-based judgments • Avoid Proxies: e. g. , Institution or Lab (i. e. , training environment) because there is bias in who can gain access to these environments. • Better: Rate candidates based on direct evidence of success (scholarly achievements) 20

3. Rubric Design • Examples • Most biased: Rank candidates; Score candidates (5= excellent, 3. Rubric Design • Examples • Most biased: Rank candidates; Score candidates (5= excellent, 1=poor) • Least biased: Score candidates on individual qualifications. Decide ahead of time how you will know it when you see it. Think broadly and inclusively about how candidates can show evidence of qualifications. 21

3. Rubric Design • How will you know it when you see it? • 3. Rubric Design • How will you know it when you see it? • Example: Research Achievements • Example: Contributions to Diversity 22

3. Rubric Design • Avoid changing scores if the result does not match your 3. Rubric Design • Avoid changing scores if the result does not match your “gut feeling” • Avoid shifting criteria • • We shift our criteria when we prefer a candidate Study of review of job candidates for stereotypically male job requiring both education & experience: • when gender not given: 76% of reviewers chose more educated candidate; 48% said education is top criterion • when male more educated: same result • when female more educated: 43% of reviewers chose more educated candidate; 22% said education is top criterion Norton, Vandello, & Darley, J. of Personality and Social Psych. 87, 817 -831 (2004). 23

4. Recruit Actively • Women may self-select out of applicant pools. Hopkins 2006 MIT 4. Recruit Actively • Women may self-select out of applicant pools. Hopkins 2006 MIT Faculty News Letter • Encouragement works! In randomized controlled trial, women responded strongly to encouragement to apply (to graduate school in applied statistics). Unkovic et al. 2016 PLo. S One 11(4) e 0151714 • Diversity in applicant pool may reduce the effects of bias. 24

5. Apply a rubric at every stage 1. First pass (e. g. , 200 5. Apply a rubric at every stage 1. First pass (e. g. , 200 -300 applicants) Rubric: Fit to job description, Research excellence 2. Second pass (e. g. , 30 applicants) Rubric: Research, Contributions to diversity, Fit within department 3. Video interview (10 applicants) Rubric: Research (future potential), Teaching & Mentoring, Contributions to diversity, Fit within department 4. On campus interview (5 applicants) *Rubric: Research (future potential), Contributions to diversity, Fit within Dept. * Provide faculty at large with rubric for providing input on visiting candidates. 25

5. Apply a rubric at every stage A special note on letters of recommendation. 5. Apply a rubric at every stage A special note on letters of recommendation. • Letters are biased. Trix & Psenka (2003) Discourse & Society, Vol 14(2): 191 -220; Madera, Hebl & Martin (2009) Journal of Applied Psych. , Vol 94: 1591. • Because letters reflect opinion, there is no clear way to apply a factbased rubric to their contents. 26

Best Practices in Faculty Searches We all have the same goal: A fair process Best Practices in Faculty Searches We all have the same goal: A fair process that efficiently and objectively identifies outstanding candidates.