• Have any questions?
  • info.zbook.org@gmail.com

Chapter 8 – Implicit Bias And The Workplace

1m ago
7 Views
0 Downloads
388.73 KB
13 Pages
Last View : 3d ago
Last Download : n/a
Upload by : Rosemary Rios
Share:
Transcription

C H A P T E R8Implicit Bias and the WorkplaceO U T L I N EWhat’s in a Scientist’s Name? 128Blind Orchestral Auditions 129Double-Blind Literature Reviews 130Success and Likeability Do Not Go Hand in Hand (for Women) 131The Selective Absence of Female Conference Speakers 131The Power of Language 133The AWARDS Project 134Conclusion 137I am concerned that saying no to service work will negatively impact my career.—A man working in biological sciences, age 26–35 years, single, CanadaIn my institution, there is little opportunity to go up in rank based on scientificmerit alone. —A man working in electrical engineering, age 36–45 years,married/partnered, NetherlandsUnderlying many of the struggles women in science, technology,e ngineering, and mathematics (STEM) endure, in both advancing theircareers and finding time and energy for family life, is the issue of implicitbias. Implicit bias is also the “elephant in the room” that will have to beshoved aside if promising programs such as those described in the casestudies are ever to succeed fully. Granted, prejudice exists throughouthuman culture. The body of literature committed to understanding whythis seemingly negative quality has persisted through our evolutionEquitable Solutions for Retaining a Robust STEM 15-5.00008-7125Copyright 2014 Elsevier Inc. All rights reserved.

1268. IMPLICIT BIAS AND THE WORKPLACEinto modern times is expansive. What we do know is that we are socialcreatures, and recognizing that there is an “us” and a “them” probablyconferred serious historic advantages. The problem is that this response,although it perhaps served a function when we were transitioning fromnomadic tribes to agrarian societies, has become a hindrance to corporateand academic progress. When we make decisions based on ideas of what“us” means, we often do so informed by the culture in which we wereraised and with a strong emotional basis rather than a logical one. Unfortunately, while this tendency is pervasive, scientists and engineers oftenthink that they have evolved beyond the rest of human society and arethus immune to allowing such passion to influence their judgment. However, studies demonstrate this is not the case.Recognizing that our prejudices occur not only at a conscious level butalso an unconscious one is important for understanding how to inoculate against them. While we may think we are focusing on one immediate thing, our brain is busy making other judgments and decisions aboutother aspects of that thing. These assessments are based on the memoriesand emotions that inform our perspective. This phenomenon is best demonstrated by a video that asks the viewer to count the number of timesa ball is passed between players in white shirts. While the action occurs,other people in black shirts elsewhere onscreen are passing a differentball. Midway through the exercise, a man in a gorilla suit walks betweenthe players, beats his chest, and then walks off screen. In a classic demonstration of selective attention, although everyone who watches the videoof course sees the man in the gorilla suit when he appears, many viewerscompletely fail to notice him.As this anecdote illustrates, oftentimes, the world as we perceive it maynot be the world as it actually is. A different video that made the rounds onthe Internet a few years ago explained that the amount of information withwhich we are inundated daily, monthly, and yearly has been increasing atan exponential rate. Some studies estimate that at any given moment, weare exposed on average to 11 million different things in our environment;however, our brain permits us to process only about 50 things at a time.As a consequence, there is a tremendous risk that we will miss things thatdo not fit into our limited view based on our cognitive processing capacity. Part of how our brain copes with this limitation is by injunctification.70Injunctification posits that we are motivated to deem the current state ofthings we see as natural and desirable, leading to the defense of the statusquo. Thus, for example, people are likely to defend the idea that womendo not like math or engineering because they are not well represented inthe field; they consider this supposed discomfort to be the reason womendo not pursue these professions, rather than considering that there may beother barriers to women’s participation. That perception is a very seriousproblem.3. ADVANCING COMPREHENSIVE SOLUTIONS

8. IMPLICIT BIAS AND THE WORKPLACE127One argument against the status quo is that countries with less rogressive cultures have more women going into STEM professionspthan is currently seen in the United States. Women in many countries inthe M iddle East and Central Asia are earning degrees in STEM fields atunprecedented rates (although some cultures are not so progressive thatwomen are free to practice their profession once they have earned a degree).Understanding the manner in which culture influences the decisions wemake, the activities we pursue, and the beliefs we hold is important tocombating these influences. Certain ideas are associated with certain values, a m essage that is consistently reinforced in our culture, whether itis in regard to appearances, constructed gender roles, racial associations,sexual preferences, or religious beliefs. Our interests in e ngineering a better scientific workplace and implementing successful strategies is overcoming the notion that “only men do science”.A classic study asked high school students to draw a scientist. Withgreat consistency, the students depicted a white man in a laboratory coatwith glasses.71 This stereotypical response alone is a problem. An even bigger problem is that when scientists were asked to participate in the sameexercise, they tended to produce the same images as the high school students. This association—that scientists are men—is reinforced continuallythroughout the media, academia, and popular culture. To examine one’sown biases, the Implicit Association Test (IAT), developed by researchersat Harvard, allows one to test one’s own preconceptions based on howquickly one responds to various word pairings.72 Although not withoutits critics and shortcomings, the IAT is still informative and a great jumping-off point for exploring the impact subtle cultural cues have onour personal development.Understanding these biases and associations is also important becausethey have a chilling effect not just on efforts to recruit girls into scienceand engineering, but also on perceptions of women in these fields. Thesesubtle associations manifest in big ways. For example, stereotype threatis the fear of performing poorly in a particular field where our gender orrace is believed to be inferior, thus reinforcing the stereotype.73 Researchers who study stereotype threat have demonstrated that girls as young as9 years start to internalize the message that “math is hard” and that it isprobably something at which boys are better. Although these associationsbegin at a young age, they persist in our subconscious and ultimately havean impact on decisions regarding hiring, promotion, and recognition. Theensuing implications are dire for the retention of women in STEM fieldsas well as in other parts of the workforce. The remainder of this chapter summarizes studies from a variety of fields and describes a projectthat applies many of the principles gleaned from these studies. The chapter concludes with a list of best practices for minimizing the impact ofsubconscious bias.3. ADVANCING COMPREHENSIVE SOLUTIONS

1288. IMPLICIT BIAS AND THE WORKPLACEWHAT’S IN A SCIENTIST’S NAME?As mentioned above, scientists may be particularly loath to concedethat they, too, are influenced by the culture in which they were raised.Most presumably would espouse an egalitarian worldview that does notendorse an innate difference between genders in their ability to performhigh-quality research. However, beliefs do not seem to translate intounbiased evaluation of individuals, suggesting that a conscious worldview is not always sufficient to overcome implicit biases. Perhaps it isless surprising that only a handful of studies have examined the impactof unconscious bias upon hiring decisions in academia.One of the approaches frequently used in sociological research inthis area is to take a standard measure of evaluation—a resume or curriculum vitae (CV), for example—and then change one facet, such asthe gender or race of the individual to whom it belongs. The CVs orresumes, identical save for that single variable, are then handed outto individuals qualified to evaluate the professional criteria. In onestudy, a CV for a psychology professor position was being evaluatedby primary investigators in the field. In a second study, an application p ackage for a laboratory manager position was assessed. The onlyvariable that differed in each of these two studies was whether the individual being evaluated had a man’s or a woman’s fictitious nameat the top of the CV.In both instances, the “female” candidate fared worse than themale candidate by a significant margin. Perhaps more importantly,the quality of the female applicant’s resume was undervalued by boththe men and women doing the evaluation, demonstrating that it isnot just men who are biased against women in science. In the case ofthe psychology professor position, the study revealed that the evaluators more positively valued the man’s research, teaching, and servicebackground; further, it confirmed previous data indicating that searchcommittees are more likely to hire a man at the associate professorlevel—although a woman is typically deemed better suited towardan assistant professorship.74 In the application package for the laboratory manager position, the individual being evaluated was a recentgraduate with a bachelor’s degree and some prior research experience.75The “female” applying for that position was viewed as less worthy ofmentoring, less competent, less desirable as a new hire, and was offeredless financial compensation. These results are distressing, particularlythe fact that the hypothetical search committee of psychologists woulddemonstrate such partiality, given that there is a substantial likelihoodthat they would be routinely exposed to papers regarding the topic ofimplicit bias during their training and in the course of their work.3. ADVANCING COMPREHENSIVE SOLUTIONS

Blind Orchestral Auditions129While the data in these studies would seem difficult to refute,there are those who try to minimize the significance of such results.Undoubtedly most of the study participants would deny being consciously biased against women in science. However, Ben Barres, anoted neuroscientist, has written eloquently on this subject based onhis firsthand experience with gender bias and double standards, having formerly been known as Barbara Barres. In perhaps the single mosttelling anecdote, he recounts how, after a highly successful presentation at the institution where he earned his PhD, a friend told him thatafter the applause died down he heard one professor remarking toanother, “Ben Barres’s work is much better than his sister’s”. In fact, hewas actually just recalling the same scientist’s presentation from a fewyears earlier.76 The conclusion that can be drawn from these studies isthat scientists are just as biased as the rest of humanity. Acknowledgingthis reality is crucial if we are to move forward, and it should not bepermitted to interfere with evaluations that should be based on scholarly merit rather than gender.BLIND ORCHESTRAL AUDITIONSOrchestras, like the fields of science and engineering, have historicallyhad a low representation of women in their ranks. In part, this is becausemany music directors, who have historically had a significant influenceover hiring, had vocally expressed the opinion that, based on auditions,women were less talented musicians than men. The traditional progression for the musically elite (man) was to train with the best (men) andthen be recommended for auditions with the top-tier orchestras. Socialtrends in the 1970s, along with many other aspects of the culture, weremoving toward a more democratic system, one based less on groomingand social nepotism and more on pure ability. As a consequence, a substantially greater number of individuals were auditioning for orchestrasbecause openings and auditions were announced more broadly. In a further effort to ensure an unbiased audition process, it became routine forthe individual performing to be placed behind a screen so that nothingabout that individual except the performance was being evaluated. Thisintervention is presumably one reason the number of female musiciansin symphonies began to increase.However, not all musical groups opted to use the screen. Using datafrom auditions and hiring numbers, Claudia Goldin and Cecelia Rousecompared the numbers of women hired in orchestras that used a screencompared with those that did not between 1950 and 1995.77 These data3. ADVANCING COMPREHENSIVE SOLUTIONS

1308. IMPLICIT BIAS AND THE WORKPLACEallowed the researchers to determine whether bias played a role in thenumber of women hired by eight separate orchestras. Although broadening the number of opportunities to audition increased the representationof women on its own, the use of a screen increased the likelihood that awoman would be hired by roughly one-third. The data from this studydemonstrate that whenever it is possible, masking the identity of the individual being evaluated to focus specifically on the performance qualitybeing evaluated is the best policy. Additionally, the more broadly disseminated the solicitations are for a position, the greater the chance of gettingthe best quality applicants.DOUBLE-BLIND LITERATURE REVIEWSMany scientific journals perform a double-blind peer review of submitted literature, a system of evaluation in which neither the author’s northe reviewer’s identity is revealed to the other. The double-blind systemis not without its critics (the small populations in certain fields sometimesmean that the primary investigator may be guessed from the subject matter, and the process adds to the administrative burden on the publisher’sside). However, the system is superior to single-blind review in terms ofcombating gender bias related to authorship.A certain ecology journal transitioned from single-blind review to double blind in 2001. Using this opportunity to compare the rate ofpublication by each gender before and after the transition for severalyears on either side of change, A.E. Budden et al. initiated a study toinvestigate the outcomes of this transition.78 After analyzing the numbers, as well as comparing them with representation of female authorsin other major ecology publications, they found that the double-blindreview process led to a nearly 8% increase in the number of articlespublished by women relative to the total number in just a few years.Eight percent may not sound like a substantial increase, but womenare poorly represented in the upper echelons of the ecology field, sothat increase may have a larger trickle-down effect. Furthermore, if thedouble-blind review increases the equity of the journal review processby reducing the opportunity for bias and subsequently leading to anincrease in the number of women publishing, removing one obstaclefrom women’s ability to succeed is a step in the right direction. A substantial number of publications are necessary for competitive positioning when applying for faculty positions and getting tenure, as wellas for winning scholarly awards; therefore, it is only logical that alldisciplines should adopt processes that reduce bias and thus permitevaluators to focus exclusively on the quality of the science.3. ADVANCING COMPREHENSIVE SOLUTIONS

The Selective Absence of Female Conference Speakers131SUCCESS AND LIKEABILITY DO NOT GO HAND INHAND (FOR WOMEN)Heidi Roizen is a highly successful, award-winning executive, venturecapitalist, and entrepreneur. Her resume demonstrates her talents in arange of areas as well as her ability to adroitly recognize and fluidly takeon challenging new tasks while climbing the corporate ladder. StanfordProfessor Frank Flynn, PhD, developed an experiment using Roizen’s stellar resume as the framework to construct a case study to examine attitudestoward success and likeability as they relate to gender. Half of the studentsin a Harvard Business School class received a case with her name includedin a document delineating her career achievements; the other half receivedthe same case except that “Heidi” Roizen was changed to “Howard” Roizen throughout the document.79 Students were then asked to readthrough the resume and evaluate the “candidate” for various qualities.While analysis showed that students found Heidi/Howard to beequally competent, they were much more severe in judging Heidi’s personality. They did not like her and generally did not want to hire or workwith her. They were put off by her aggressive nature and found her “selfish”. The same cannot be said for their evaluations of “Howard”. Thisresult is consistent with other literature, which generally finds that success negatively correlates with likeability for women. This equation isborne out in the real world when employees are interviewed by the press;regardless of whether the workplace is the New York publishing worldor a Silicon Valley tech start-up, high-achieving women are consistentlyrepresented as odious and demanding. This negative association betweenwomen and success is a constant struggle to women’s advancement andaccounts at least partly for the lack of female leadership at technologyand pharmaceutical companies, medical schools, and Fortune 500 corporations. The one upside to the above findings is that the negative correlation between likeability and success can be reduced when high-achievingwomen develop personal relationships with their colleagues and others;research shows that men tend to judge accomplished women they knowbeyond a one-dimensional context less harshly. However, such familiaritystill does not eradicate the problem.THE SELECTIVE ABSENCE OF FEMALE CONFERENCESPEAKERSConferences are an important opportunity for individuals to present their research, disseminate their findings, receive feedback on theirresults, initiate collaborations, and network with other experts in their3. ADVANCING COMPREHENSIVE SOLUTIONS

1328. IMPLICIT BIAS AND THE WORKPLACEfield. Yet women tend to be heavily underrepresented among invitedspeakers at disciplinary conferences. The reasons for this are complexand multifaceted. There is, of course, variation among disciplines in theirrepresentation of women. One recent study showed that the likelihood ofwomen being invited to speak depends on the representation of womenand men on conference organizing committees. In several subdisciplinesof primatology, for example, women have comprised more than half thefield for the last few decades. However, when the study authors examinedthe number of invited female speakers who participated in symposia overthe past two decades (separate from those who requested the opportunityto present their work), they found that the composition of the selectioncommittee had a tremendous impact on the number of women speakersinvited. All-female and mixed-composition committees invited womenspeakers at a rate commensurate with the representation of establishedleaders in the field. An all-male committee, on the other hand, invited20–35% fewer women than would be expected as symposium speakers.The authors propose that the gender inequity seen with the all-male committee is due most likely to subconscious bias or homophily (the tendencyto want to bond with one’s own kind), a term that presumably comes upwith some frequency in primatology.80One aspect of this study that was not addressed concerned how manywomen were invited to speak compared with how many accepted theoffers. Another recent study indicated that in some disciplines, althoughthe number of women invited is representative of the populations withinthe discipline or hosting society, many women decline the opportunityto speak.81 This reluctance could certainly be a very valid explanationfor the low representation of women in conference symposia in the fieldbeing examined—evolutionary biology—a discipline in which women aresignificantly more poorly represented than in primatology. However, thereasons are perhaps more complicated than simply that women do notwant to speak on the record. Women still face greater scrutiny online, during interviews, via applications, by review panels, and in all other aspectsof their scientific careers due to implicit bias. So, although women mightbe turning down an opportunity to speak, thus declining the chance topromote their work and serve as a role model, at the end of the day theimplicit bias that does not associate women with science—or rather, thatassociates science with men—may be a large part of the underlying sourceof the discrepancy. It would be interesting to see whether the invitationcoming from an all-female versus all-male selection committee had anyimpact on the acceptance rate, and whether there is a tipping point for arepresentation level in a particular field that would increase the likelihoodof more women accepting invitations.The key lesson of both these studies is that bias may play a role in therepresentation of women at conferences, which is one more way that3. ADVANCING COMPREHENSIVE SOLUTIONS

The Power of Language133women wind up appearing marginalized in the STEM fields and thuspublically presenting fewer positive role models and limiting their owncareer networks. In addition to actively encouraging greater participationby women at conferences, an approach designed to minimize implicit biasshould include creating heterogeneous committees. Furthermore, assuring all participants have access to the financial and family-care accommodations necessary to attend conferences is a step in the right directiontoward increasing the representation of women in symposia.THE POWER OF LANGUAGEWords have the power, subtly or overtly, to shape perspectives andcolor the accomplishments listed on a person’s CV or resume. The tintsused can impact dramatically how natural the final image appears. Coloring a picture of a zebra with green and purple stripes might make foran interesting image, but it is not a very realistic depiction. In the samemanner, using a certain vocabulary to describe women and another setof words to describe men can shift how realistically one might view anindividual’s fit for a particular job. Many studies have demonstrated thatwhen individuals are crafting letters of recommendation for women, theychoose descriptive terms different from those they use for men.When writing about women, individuals tend to focus on the candidate’s ability to nurture, her dependability, her capacity as a team player,and perhaps her ability to balance her obligations as a mother as well asa scientist. They use words such as “conscientious”, “methodical”, and“dependable”.82 On the other hand, words used to describe men focuson their intellect and scientific achievements, including terms such as“brilliant”, “analytical”, “talented”, and “results”. Consequently, eventhough both of two letters might be very positive, each paints a verydifferent picture of the candidate’s relative ability to contribute to the institution—and thus, how he or she is viewed in terms of fit for a particularposition.83 As a consequence, considering the language being used whencrafting a recommendation letter is very important. Using gender-neutralor even conventionally masculine terms, to describe a woman is advisable.T. Schmader has compiled a complete list of recommended terms.82In addition to its impact upon letters of recommendation, certain language in a job description can bias a prospective hire. For example, ajob advertisement that emphasizes company culture—e.g., a tech start-upthat sounds more like a fraternity house than a workplace—is likely to“turn off” an applicant who is not interested in walking into an officethat is borderline hostile to women. Along the same lines, when readinga solicitation for a particular award named after a pioneer in the field,someone considering candidates for that award might be swayed by the3. ADVANCING COMPREHENSIVE SOLUTIONS

1348. IMPLICIT BIAS AND THE WORKPLACEsubconscious influence of the name of the award to nominate someonewho seems to fit the standard to receive that accolade because he or sheresembles the pioneer. Giving consideration to the language used in writing recommendation letters and in creating positions, endowments, andawards; and refraining from using names that indicate some particularexample that no longer resembles the potential breadth and diversityof the STEM enterprise would be positive steps toward creating a moreinclusive academic community that focuses on actual achievements, notstereotypes.THE AWARDS PROJECTDrawing on many of the principles outlined within this chapter, whichemphasize focusing on achievement by increasing transparency and minimizing opportunities for bias, it may be instructive to review a projectundertaken by the Association of Women in Science (AWIS) regardingrecognition. Scholarly awards from disciplinary societies are an importantbenchmark of achievement in academic culture. They are used in evaluating candidates for hiring, tenure, and promotions. Workplace management studies have shown that people would rather be recognized for theirachievements than given raises, and this finding is particularly salient inthe academic environment, which does not attract people for the moneyas much as for the opportunity to make outstanding intellectual contributions to the body of scientific knowledge.Anecdotally, it has been said that women were underrepresented amongwinners of scholarly awards given by disciplinary societies relative to theavailable pool of women holding positions in the upper echelons of academia and industry. The late Phoebe S. Leboy, PhD, a past AWIS presidentand professor at the University of Pennsylvania, examined a wide rangeof scientific societies and found this characterization to be accurate, asshown in Figure 8.21. Additionally, women were more likely to be societypresident than winners of a scholarly award. Furthermore, women wereoverrepresented among winners of mentoring and service awards relativeto the expected pool, and women won particularly few scholarly awardswhen there were “women only” awards available.84Recognizing that this disparity is a serious concern for the promotionand retention of women in STEM, AWIS sought and was awarded a grantto study ways to increase the transparency and equity of the awards process.85 The resulting project, Advancing Ways of Awarding Recognition inDisciplinary Societies (AWARDS), supported a workshop in 2010 to whichseven scientific societies representing a wide range of STEM disciplines86were invited. At the workshop, various aspects of the awards processwere examined.3. ADVANCING COMPREHENSIVE SOLUTIONS

The AWARDS Project135FIGURE 8.21 Scholarly awards to women by professional societies.The recommendations to increase the transparency and thereby theequity of the awards process involved a wide range of approaches andtopics. Central was the idea that implicit bias influences many different aspects of the process. Implicit bias training at the beginning of theworkshop was central to the discussion; in this way, all participants werehelped both to understand that everyone makes subconscious decisionsthat influence their judgment and to help inoculate against their influence.The remainder of the discussion included ways to minimize biasthroughout the awards process. A key part of getting better representationof women among award winners is by increasing the pool of candidates.For example, some awards are given annually in fields with dwindlingnumbers of participants. Thus, it is important to rethink the timing ofawards to generate larger pools from which to select the best nominees.When the nominations open, it is imperative to raise awareness by crafting solicitations that use gender-neutral language and then distributingthem broadly through a variety of media, both print and online. Updatingthe organization’s Web site to consolidate the location of the awards andtheir criteria on a single page, so that they are easy to find, is also encouraged. Creating canvassing committees and reaching out to women’sand minority groups is another way to broaden the pool of nominees.Nomination and selection committees should also reflect the breadthand depth of the society’s membership by ensuring gender and ethnicdiversity. When these committees convene, they should begin by inoculating against implicit bias, discussing which specific qualities are to be3. ADVANCING COMPREHENSIVE SOLUTIONS

1368. IMPLICIT BIAS AND THE WORKPLACEevaluated, and disclosing any conflicts of interest so that the caliber of thenominees’ credentials is what is being evaluated rather than the strengthof their professional networks.At the workshop’s conclusion, each society developed its own list ofbest practices to suit the nature of its own discipline. These recommendations were taken back to the society for acceptance and ratification. Thesocieties then implemented the changes accordingly and shared news ofthe project via their respective journals and member publications. AWIStracked the progress of their awards program over the next 2 years andheld a follow-up workshop in 2012 that included the 7 original pioneersocieties as well as 11 new societies that approached AWIS after hearingabout the project.The outcome of the AWARDS project’s first two years in regard to itsintended goal of increased representation of w

ball. Midway through the exercise, a man in a gorilla suit walks between the players, beats his chest, and then walks off screen. In a classic demon- . this area is to take a standard measure of evaluation—a resume or cur - riculum vitae (CV), for example—and then change one facet, su