In this article, we present the development and validation of the Perceived Subtle Gender Bias Index. Given the inherent difficulty in identifying and measuring the perceptions of subtle gender biases, this index provides researchers and interventionists with a tool that does not require participants to identify/label an event as a gender bias incident. We used a mixed method and constructivist approach that prioritized and privileged the voices and experiences of women in science, technology, engineering, mathematics (STEM). The current article describes two studies: (1) index development and (2) index refinement and validation, using a national survey of women academics (N ¼ 882). Findings support a four-subscale structure, including perceived gender inequity, collegiality, mentorship, and institutional support. Methods and analyses support face, convergent, discriminant, and predictive validity for the use of the index among academic faculty women. Additional online materials for this article are available on PWQ’s website at http://journals.sagepub.com/doi/suppl/10.1177/0361684319877199
subtle gender bias, sexism, microaggression, micro-inequity, workplace climate, academia, scale development, faculty women
There is a growing awareness that gender biases persist in academia. Researchers have reported that these biases contribute to negative consequences for women, relative to men, including a slower pace of career advancement (McBrier, 2003; Valian, 2005), unfair hiring practices (Madera, Hebl, & Martin, 2009), and women feeling a lower sense of belonging in, and satisfaction with, their academic workplace (Moors, Malley, & Stewart, 2014). These gender differences have prompted many interventions aimed at increasing the recruitment, retention, and promotion of women (Aguirre, 2000; Bilimoria, Joy, & Liang, 2008; Stepan-Norris & Kerrissey, 2016). Yet no tool that assesses perceived and subtle gender bias, specifically in the academic context, currently exists.
The academic career can be difficult for anyone pursuing it, but particularly for women (vs. men), and especially women in STEM (science, technology, engineering, mathematics) disciplines (vs. women in other fields; Nelson Laird, Sullivan, Zimmerman, & McCormick, 2011; Xu, 2008). Bailyn (2003) described the academic career as one filled with unique challenges, such as the burden of being an expert
1 Department of Counseling & School Psychology, San Diego State University, CA, USA
2 Department of Psychiatry, Virginia Commonwealth University, Richmond, VA, USA
3 Department of Psychology, University of Massachusetts Lowell, MA, USA
4 Department of Medicine, University of Massachusetts Worcester Medical School, MA, USA
5 Department of Mechanical Engineering, University of Massachusetts Lowell, MA, USA
6 Department of Population & Quantitative Health Sciences, University of Massachusetts Worcester Medical School, MA, USA
7 Department of Psychology, University of Massachusetts Lowell, MA, USA
8 Department of Sociology, University of Massachusetts Lowell, MA, USA
9 Department of Engineering Technology, University of Massachusetts Lowell, MA, USA
10 Department of Medical Social Sciences, Northwestern University, Evanston, IL, USA
11 Department of Psychiatry, University of Massachusetts Worcester Medical School, MA, USA
Nellie Tran, Department of Counseling & School Psychology, San Diego State University, 5500 Campanile Dr., San Diego, CA 92182, USA. Email: ntran@mail.sdsu.edubeing regularly overloaded by long-term projects, and the difficulty of balancing multiple roles and expectations, including research, teaching, and service. Perhaps unlike other workplaces, the academy has standards originally developed for men, and now requiring women to live and work in a male-normed style that may create further inequitable workplace conditions (Bagilhole, 2002; Yoder, 2018). For example, Bailyn (2003) stated,
the ideal, the perfect academic is someone who gives total priority to work and has no outside interests and responsibilities. ... And in some ways that may be the greatest inequity of all: the profession is set up in such a way that men academics routinely have families, while women, given current rules, find it much more difficult (p. 139).
Yoder (2018) also highlights the masculinized academic hierarchy that privileges research/scholarship over teaching/mentorship. It is within this unique setting that we developed and assessed the validity of the current index.
Perceived subtle gender biases are perceptions that an individual holds regarding subtle experiences that are related to their gender identification. Other related concepts include microaggressions and implicit bias. These terms are used to describe subtle bias perpetrated—often unintentionally, and sometimes by well-meaning individuals—through verbal or nonverbal, visual, or environmental indignities (Rowe, 1990; Russell, 1998; Solorzano, Ceja, & Yosso, 2000; Sue, Bucceri, Lin, Nadal, & Torino, 2007).
Subtle gender biases are, by definition, hard to discern. What makes them difficult to prevent and stop is the often invisible nature of the phenomena. Unfortunately, these subtle biases often result in inequities (Rowe, 1990). Thus, perceptions of subtle gender biases are dangerous because they may lead to other psychological effects, such as stereotype threat when an elicited stereotype, regardless of intention, re-produces existing inequities and biases (Steele, 2011), and social identity threat when an individual’s identity as a woman is undermined (Cheryan, Meltzoff, & Kim, 2011). A meta-analysis of stereotype threat found that the subtler the cue of a person’s underprivileged status, the stronger the potential effect of the internal threat on projected outcomes, such as performance (Nguyen & Ryan, 2008). For example, if women enter a room showcasing photos of only men past presidents of their institution, women may experience an internal threat that triggers more significant deficits in performance, when they learn that the university had not had a woman president. This subtlety makes it possible, and even likely, that the environmental gender bias indicated by the photos goes unnoticed by either or both parties (i.e., women and those unwittingly imparting the bias through displaying the portraits). Even when women can identify the subtle biases in their experiences, they may not perceive these incidents as biases because mentally coding the incident as favorable might serve as a self-protective mechanism that women may use within hostile environments (Glick & Fiske, 2001). Therefore, measures of perceived subtle biases that rely on self-reported experiences of biases in which women must categorize, or actively classify, an incident as an experience of bias, may fail to capture the totality of women’s realities.
The workplace is one context where subtle gender biases are found and may have deleterious effects (Agars, 2004; Betz & Fitzgerald, 1987; Heilman, 2012). And the academic work setting is one workplace ripe with subtle gender biases (Bond, Punnett, Pyle, Cazeca, & Cooperman, 2004). University women faculty and students report facing sexual harassment, whereas men faculty and students report almost no experiences of sexual harassment (Fitzgerald et al., 1988). Men and women also may evaluate similar workplace interactions differently, especially around gender biases. For example, women are more likely than men to detect discrimination, especially as the bias becomes subtler (Basford, Offermann, & Behrend, 2014). Studies also show that men and women differ in supports that lead to higher levels of workplace satisfaction. Previous studies showed that women reported higher satisfaction with their academic workplace when they had more social contact with their department colleagues (Pfeffer & Langton, 1993) and when women felt their colleagues were more committed to their professional success (Trower & Bleak, 2004). For these reasons, it is important to develop tools that measure perceptions of subtle gender bias. Such an index, or measure, should focus on women’s perceptions of potential gender bias incidents and structures.
In past studies of subtle gender bias, researchers included items from several different measures/indexes. For example, Settles, Cortina, Stewart, and Malley (2007) considered a combination of two measures to understand workplace climate: sexist climate and the general negative climate. Sexist climate reflected perceptions of a department characterized by inequality between men and women. Researchers combined two items from Riger, Stokes, Raja, and Sullivan (1997, as cited in Settles, Cortina, Stewart, & Malley, 2007) and seven items from the Gender Fairness Environment Scale (Hostler & Gressard, 1993, as cited in Settles et al., 2007). In another study, Settles, Cortina, Buchanan, and Miner (2013) considered gender-based mistreatment in the workplace more broadly and employed measures of formal gender discrimination, gender derogation, and organizational sexism. Given the lack of clear measures of perceived subtle gender bias specific to the academy, we sought to develop a comprehensive measure, which we call an index, specifically for use in such a setting.
Few studies report the perceptions of subtle gender biases at different academic ranks, and those that do, are mixed in their findings. For example, women at the associate level have reported barriers that affect their ability to obtain full professor status (Buch, Huet, Rorrer, & Roberson, 2011). Given the longer time spent in the academy, tenured women, especially full professors, may report perceiving more subtle gender biases, relative to assistant and associate professors, including differential requirements for women that require them to work harder and longer (Toren, 1991), compared with men. A study of Black women full professors highlighted the many biased experiences they endured in genderized and racialized departments when attempting to achieve promotion to full (Croom & Patton, 2011). As women are promoted through the academic ranks, they are also subjected to a greater gap in race and gender representation (West & Curtis, 2006). Studies consistently report lower levels of satisfaction by academic women of color, relative to their White counterparts (Aguirre, 2000; Hagedorn, 2000; Seifert & Umbach, 2008). The publication of the book Presumed Incompetent (Gutie´rrez y Muhs, Niemann, Gonza´lez, & Harris, 2012) provided personal narratives written by women of color who detailed their experiences of overt and subtle discrimination within academia. Therefore, using a measure of perceived subtle gender biases, we may reveal more reported perceptions of bias by women of color relative to White women faculty.
The differences between non-tenure track and tenured/ tenure-track faculty are also complicated. Past research suggests that working conditions are often not ideal for nontenure-track faculty (Kezar & Sam, 2010). Researchers have reported that most institutions do not have systematic procedures for hiring non-tenure-track faculty (Cross & Goldenberg, 2009; Gappa & Leslie, 1993), leaving them vulnerable to biases within this process. Tenured and tenure-track women were also exposed to biases based on the tenure and promotion evaluation process (see Weisshaar, 2017). These differences lead to a more general question regarding the relations between perceived biases and tenure status.
Some studies report differences between women working within STEM disciplines, relative to non-STEM disciplines (cf. Nelson et al., 2011; Xu, 2008). However, in one study (Moors et al., 2014), both men and women in STEM and nonSTEM disciplines experienced dissatisfaction with their workplace when they did not feel support from colleagues. The same study found that women were especially dissatisfied when they lacked institutional support for their family choices. STEM fields have been studied for their hostile environments for women, who are disproportionality represented (see Bailyn, 2003; Bilimoria et al., 2008). It may be, that the gender composition, and not the discipline itself, is important to consider. One study reported that both men and women working in spaces dominated by men reported more institutional and interpersonal sexist experiences, compared with employees working in spaces dominated by women, and spaces that are more gender balanced (Bond et al., 2004). Therefore, using a measure of perceived subtle gender bias, we might also find that women working in men-dominated spaces will report more biases.
Subtle gender biases exist beyond policy and institutional infrastructure and include women’s perceptions. The assumptions that men and women work in the same ways, have the same workplace values—and thus require the same resources—are likely to disadvantage women (Mottazl, 1986). Given women’s different socialization and experiences, women academics may perceive workplace events and policies differently. For example, expansion of parental leave policies, to include both men and women, has resulted in increased numbers of both men and women taking leave (Han, Ruhm, & Waldfogel, 2009). Paid family leave time has resulted in higher numbers of mothers taking leave and facing an increased number of work hours upon their return (RossinSlater, Ruhm, & Waldfogel, 2013). However, men who took parental leave often ended up doing less childcare than their women counterparts and used the time to catch up on their research and writing (Rhoads, 2005). Consequently, the availability of institutional policies to support women may not be sufficient in workplaces where women do not perceive it safe to take leave (Mason, Wolfinger, & Goulden, 2013). These gender-neutral policies have also been shown to result in larger gender gaps in tenure rates as women’s tenure rates decline (Antecol, Bedard, & Stearns, 2018). Thus, a good measure of perceived subtle gender biases in the academy must include perceptions of infrastructure, knowledge, and the ability to use policies
Our objective was to develop and validate the Perceived Subtle Gender Bias Index (PSGBI) for women in academia. We used a mixed-methods approach to develop the PSGBI. In Study 1, we used a constructivist approach that prioritized the voices and experiences of academic STEM women. Using one-on-one interviews with 19 women academics in STEM disciplines across two secondary education campuses in New England, United States, we developed an initial 175 potential items for the Index. Consultants, expert in understanding women in STEM and women in the workplace, then reviewed the PSGBI. Study 2 involved a quantitative survey including the PSGBI, demographics, and three validation measures. Study 2 included responses from 882 women academics nationally and across academic disciplines, and we developed the final Index using exploratory and confirmatory factor analyses. We hypothesized that PSGBI would predict perceptions of workplace incivility, a weak negative correlation with PSGBI scores and endorsed hostile sexism beliefs (discriminant validity), a weak positive correlation of PSGBI scores with endorsed benevolent sexism beliefs (discriminant validity), and a weak positive correlation of PSGBI with reported sexist events (discriminant validity). Finally, we also tested for concurrent validity by considering the PSGBI’s ability to detect expected differences in women’s experiences. We hypothesized that women full professors would perceive more biases, compared to assistant and associate professors. We also hypothesized that women would perceive more biases if working in STEM disciplines, versus those in non-STEM disciplines, and more biases for women in predominantly male contexts (departments and fields) versus those women in contexts with fewer men. We also hypothesized that White women would report fewer biases relative to other racial and ethnic minority women. And last, we explored the relation between perceptions of bias among tenure statuses (non-tenure track, tenure track, and tenured).
We used purposeful sampling to obtain a sample of 19 women academics from STEM disciplines. Nine women were from Institution A and 10 women were from Institution B. At the time, Institution A had 203 and Institution B had 129 eligible women STEM faculty. Eligibility criteria were: (1) a current faculty affiliation at either institution in an STEM department, (2) a PhD or equivalent degree, and (3) appointment as a tenured or tenure-track faculty member at their respective institution for at least 6 months. STEM included science, technology, engineering, and math disciplines. Based on local institutional culture, some disciplines were excluded (psychology, sociology, anthropology). Participants were six assistant professors, six associate professors, and seven full professors. Despite multiple recruitment strategies and attempts, the study included only one faculty woman of color, perhaps due to base rate issues and concerns about protecting confidentiality and anonymity scores above and beyond ASI and gender ratios within a small cohort. Some STEM departments included only one eligible woman faculty member, and our interview team followed both Institutional Review Boards’ protocols and did not collect or report individual faculty department information; we thoroughly de-identified interview transcripts before the analysis team received access to the files.
Two interviewers unaffiliated with either campus, with extensive experience with conducting in-depth interviews of sensitive topics, conducted all interviews. Interviews lasted approximately 60 minutes. Participants received a $25 gift card as compensation. Interview questions focused on participants’ experiences with their respective field throughout their careers (past and current), with a focus on educational, job-seeking, career advancement, and mentoring experiences. Refer to the Supplemental Materials at http://journals.sagepub.com/doi/suppl/10. 1177/0361684319877199 for the full interview guide. Interviews were recorded and transcribed verbatim by a third-party company. We used a three-level deidentification process including the transcriber, the interviewer, and the interviewee. First, the transcription service removed all information that might identify the participant including geographical locations, program names, and research field identifiers. Second, the interviewer read each transcription to ensure accuracy and proper deidentification of data. Last, the interviewee had the option to review the transcriptions prior to their release to the research team. No administrators from either campus had access to the raw interview data.
Qualitative content analysis (Neuendorf, 2002) of the transcribed interviews involved five researchers on the analysis team: four academic research psychologists (two assistant professors, one associate, and one full professor) and one graduate research assistant. First, all five researchers coded five 60-minute interviews and reviewed each code as a team to develop a clear team understanding and coding manual for perceived subtle gender biases. Then, each interview was individually coded by two researchers. Researchers discussed codes until 100% agreement was achieved for each interview. We operationalized subtle biases based on Swim and Cohen’s (1997) definition, that overt and blatant sexism refers to harmful and unfair treatment of women that is intentional, visible, and unambiguous. Subtle or covert sexism is hidden or unnoticed because it is built into cultural and societal norms and might be unintentional. Subtle biases referred to treatment, behaviors, perceptions, thoughts, and attitudes that might be different for men and women, and included helping behaviors and objectification—women feeling that their bodies might be more important than their intellectual work. Most important, coders looked for instances when women mentioned an incident as creating a differential outcome for her, relative to other colleagues. This difference could be either positive or negative.
The following section describes topics women shared in their interviews. The subtle nature of these biases made them difficult to describe outside the context of interview transcripts. Often, women discussed how their careers “felt different” from their male colleagues, how they felt disadvantaged, or how their experiences as women made things different for them in both positive and possibly harmful ways. Coders discussed all discrepant codes until we achieved full consensus. Qualitative content analyses revealed subtle biases pertaining to six topics: (1) general attitudes, (2) visibility of women and their work, (3) access and allocation of resources (formal and informal), (4) mentoring, (5) clarity of and assistance with promotion and tenure, and (6) hostile cultural climates. Below are brief descriptions of each theme. Ideas within these themes guided our writing of items.
Participants often spoke about gendered attitudes and beliefs that they or their colleagues held. These differed from the other categories because they were beliefs and attitudes, not behavioral categories. For example, women reported perceived attitudes that women needed to be on campus and visible to colleagues, in order to be seen as working; generational differences about resources that younger or newer colleagues received, relative to more advanced colleagues; that women’s careers are enhanced by having a high-powered STEM male faculty partner; preferred personality types; the suitability of the academic lifestyle for women, especially those that had children; and women lacked the desire to obtain a doctorate due to the length of time required. While for both men and women grant funding is tied to resrouces, and men and women might receive the same number of grants, women were often granted less money by the granting institution. Women also discussed the extent to which colleagues held subtle and blatantly sexist beliefs and that many of their colleagues were not aware of these attitudes being discriminatory.
Participants spoke about the various ways that they received, and did not receive, visibility for themselves and their work. They also talked about individuals who enhanced their visibility. For example, one participant reported a mentor working to make sure she held positions on a project that would be most visible to grant program officers. Several women indicated the importance of being seen physically at work. As mentioned earlier, women also reported attitudes about visibility that resulted in their feeling, and sometimes being told, that they were not working unless their work happened on campus where men and more senior colleagues could see them.
Participants discussed the extent to which they had access to resources and how institutions distributed resources. These resources included differential salaries, space allocations, leave policies that disadvantaged women, daycare options, role models, professional development workshops, and availability of onboarding information (i.e., orientations, introductions).
Participants discussed the various forms of support they received from peers and colleagues at their institutions. Many participants also reported missed opportunities for mentoring relationships and the lack of availability of mentorships. Mentors supported faculty by providing personal and professional advice and sponsoring their work.
In Study 2, we statistically reduced, refined, and evaluated the validity of the PGSBI, in an original sample of U.S. women-identified academics. The interview and coding process in Study 1 led the research team to develop an item for each coded subtle bias event, 175 total potential items. Similar to the coding process, in which codes were combined to create prevalent topics, we combined similar items and created a list of survey items that best represented the interviewee’s experiences of bias. This refinement and reduction process resulted in 122 unique survey items and experts added 2 additional items. The final 124 items represented both positive and negative experiences that led to differential experiences (See online supplemental file at http://journals. sagepub.com/doi/suppl/10.1177/0361684319877199).
Next, the 122 items were reviewed by 10 experts who came from both institutions, which resulted in the addition of 2 items for a total of 124 total items. These experts held specialized knowledge on gender in the workplace (n ¼ 5), scale development (n ¼ 1), and STEM academic women (n ¼ 4). They reviewed items to ensure they matched the experiences women had in their workplace, the current scholarly literature, and fit survey instrument creation standards.
We recruited participants through academic listservs (e.g., Divisions of the American Psychological Association, Society of Women Engineers, National Science Foundation ADVANCE), personal networks of those on the research team, social media outlets (e.g., Facebook, Twitter, Blogs), and flyers at the 2015 Academic Women in Science (AWIS)/National Science Foundation-ADVANCE (NSFADVANCE) conference. Participants had the opportunity to enter a raffle to win one of twelve $250 gift cards. The survey was administered electronically via Qualtrics.com to academic women across the U.S. Participants were included in the analyses if they identified as a woman and currently worked in a U.S. academic institution in a faculty position.
The initial sample included 1,353 participants. Analyses, however, excluded the following 471 surveys: 122 were missing gender identification, 95 were men (although we initially elected to include men in our survey, men participants reported that items did not pertain to them and our interest was in developing a measure based on women’s experience; we excluded all men), 16 were missing U.S./ international status, 14 were faculty working internationally, 149 were lecturers, 47 were adjunct faculty, and 28 were other faculty. The remaining 882 women were assistant, associate, or full professors in the United States, 779 of whom provided complete data on items included in our factor analyses. As presented in Table 1, the analytic sample (N ¼ 882) was close to evenly split among the three faculty ranks, and over half had tenure at the time of survey completion. Most reported ages between 31 and 50 years, and almost 40% had been at their current institution for more than 10 years.
The instruments used in the online Qualtric’s survey administered nationally to women academics across disciplines included demographics, a measure of workplace civility, hostile and benevolent sexism, sexist events, and perceptions of subtle gender bias. A demographic questionnaire included personal (gender) and workplace demographics (tenure status, academic rank, field/discipline, age, time employed at workplace). Refer to Table 1 for details.
To evaluate the 21-item PSGBI, we first analyzed the missing data using a split sample exploratory factor analysis (EFA), followed by the second split sample confirmatory factor analysis (CFA), and analyses to test the index for concurrent, discriminant, and predictive validity with the current sample. We evaluated concurrent validity with the following hypotheses: (1) White women would perceive significantly fewer biases relative to other racial and ethnic minority women. (2) The PSGBI would differentiate the reported level of PSGBI experienced by academic status such that full professors would report perceiving more biases than associate and assistant professor women. (3) On an exploratory basis, we hypothesized that tenured women would report more subtle gender biases relative to non-tenured women or tenure track women. We also hypothesized that (4) women in the STEM disciplines would perceive significantly more subtle gender biases relative to women in non-STEM disciplines. (5) We hypothesized that women in men-dominated contexts (departments and fields) would perceive significantly more subtle gender biases relative to settings with fewer men. We also hypothesized (6) convergent validity such that women would report higher PSGBI scores, and more sexist events (SSE, Matteson & Moradi, 2005). We hypothesized discriminant validity using (7) hostile sexism beliefs, (8) benevolent sexism beliefs (ASI; Glick & Fiske, 1997), such that PSGBI’s would be negatively related to hostile sexism beliefs and positively related to benevolent sexism beliefs. These relations would also be small (Cohen, 1992).
Following Lindner and Tantleff-Dunn (2017), we conducted exploratory factor analysis (EFA) using multiple stages, rerunning EFA to prune items until all retained items met criteria, which included (a) a primary factor loading of 0.55, considered at least “good” (Tabachnick & Fidell, 2011); (b) loadings on other factors were at least .20 smaller than the primary factor loading (Costello & Osborne, 2005; Herbozo & Thompson, 2006); and (c) communality 0.4 (Lindner & Tantleff-Dunn, 2017). We used varimax rotation.
Using the second half-sample, CFA was conducted to confirm the factor model indicated by the EFA. As outlined by Browne and Cudeck (1992), model fit for the CFA was assessed using w2 and degrees of freedom (values <3 indicate good fit), root mean square error of approximation (RMSEA, values <0.06), Bentler’s comparative fit index (CFI, values 0.95 indicate good fit), standardized root mean square residual (SRMR; values <0.05 indicate good fit). Fit statistics for CFA on the second half-sample for the three-factor structure indicated by the EFA included w2 /df ¼ 2.47, RMSEA ¼ .0580, 95% CI [.0499, .0661], CFI ¼ .9470, SRMR ¼ .0453. With the exception of the CFI, all indexes reflect a good model fit.
Overall, within the current sample, .75% of item values (834 of 882 participants on each of the 124 items) were missing; 8 of the 124 items had observed values for all 882 participants. All items had no more than 5.8% missing values. Of the 882 participants, 80.1% (N ¼ 706) had complete data on all 124 items. Once reduced to 21 items, the missing values report was 0.45% of item values (83 of 882 participants by 21 items) were missing; 2 of the 21 items had observed values for all 882 participants. All items had no more than 1.6% missing values. Of the 882 participants, 94.6% had complete data on all 21 items. Last, we compared CFA fit for STEM and nonSTEM women (including both random half-samples) to ensure adequate fit. For STEM: w2 /df ¼ 2.53, RMSEA ¼ .0639 (.0552, .0726), SRMR ¼ .0504, CFA ¼ .9363. For non-STEM: w2 /df ¼ 2.39, RMSEA ¼ .0557 (.0476, .0638), SRMR ¼ .0416, CFA ¼ .9530. This suggests adequate fit for both samples.
To test for convergent validity, we computed totals for each of the four subscales and the overall total of the 21-item index. We correlated these scores with similar scales, listed below, to assess validity. Because these scales were rightskewed, we calculated Spearman correlations (see Table 4). Table 1 summarizes means and standard error for race/ethnicity, academic rank, tenure status, discipline, gender ratio in department, and gender ratio in field for each PSGBI factor, using analysis of variance (ANOVA) to compare groups. To account for multiple hypotheses testing across the 24 ANOVAs, we applied a Bonferroni correction to each omnibus F-test.
We developed and tested the validity of the Perceived Subtle Gender Bias Index (PSGBI) within our sample, as a distinctive measure of women’s perceptions of subtle biases in their current academic workplace. We found evidence of convergent validity in our sample, as determined by a significant positive correlation between perceived subtle gender biases with both workplace incivility and reports of sexist events. We also showed discriminant validity within this sample from measures of sexist attitudes (hostile and benevolent sexism) showing a small effect. We found PSGBI factors of Gender Inequality (Factor 1) and Institutional Support (Factor 4) showed a small negative relation to hostile sexist beliefs. Only the Collegiality factor (Factor 2) showed a small positive relation to benevolent sexist beliefs. These findings suggest the value of the PSGBI aligns with another established measure of subtle workplace biases and reports of sexist experiences and yet is different from other more overt measure of hostile sexist attitude. Sexist beliefs, as measured by the ASI, have often been discussed in relation to the workplace but have primarily focused on reactions to vignettes about the workplace, rather than actual experiences in the workplace (e.g., Dardenne, Dumont, & Bollier, 2007; Glick, Diebold, Bailey-Werner, & Zhu, 1997). The failure to find relations between the benevolent sexism measure and three of the PSGBI subscales suggests that women’s beliefs about benevolent sexism may only correspond to women’s perceptions of biases with colleagues, given that benevolent sexism tends to be relational. On the other hand, if benevolent sexism exists within relationships, this may indicate possible missing items in the PGSBI mentorship factor.
Scholars suggest that the psychological effect of microaggressions, implicit bias, and other forms of subtle gender biases is not additive, but rather exponential (Nadal, 2010). Thus, the need to explore this aspect of microaggressions and the resulting micro-inequities in a longitudinal manner is evident, though this and prior studies have yet to follow a cohort over time. We recommend that in the future, researchers use the PSGBI to assess academic settings over time, to understand the effect of intersectional identities and experiences, such as those of women of color more specifically. Scholars might consider the effects of perceived subtle gender biases directly and indirectly on other psychological wellness and physical health outcomes. And the PSGBI could be used to assess whether the intensity of perceived subtle gender biases affects career-related outcomes, such as publication productivity, faculty evaluations, and the number of institutions where individuals have worked.
The PSGBI was developed with the primary purpose of better assessing the perception of the experiences of women in science, technology, engineering, mathematics (STEM) (Study 1). Therefore, the perceptions of women in other work disciplines and settings are likely missing from this index. We recommend that institutions conduct a self-study in order to better understand the perceptions of subtle gender bias in their organization and design interventions to best meet their institution’s needs. Researchers and administrators may consider adapting the index as necessary to capture perceptions of domains and experiences most relevant to their contexts
The PSGBI serves as a valuable tool for understanding the perceptions of women faculty within the academic workplace. Based on the qualitative techniques used to develop the items for the measure, the PSGBI also provides a helpful tool for promoting awareness of areas and experiences that may lead to perceptions of differential treatment (i.e., bias). The items of the PSGBI may not reflect experiences that are entirely unique to women academics; however, they are linked to perceptions of differential treatment and outcomes. The process detailed in this article also provides a useful strategy for institutions and other workplaces to conduct effective self-studies that are rooted in the voices of marginalized peoples within that context. We especially encourage those in the position of selecting measurement tools (e.g., department chairs, diversity and equity coordinators, National Science Foundation ADVANCE programs) to consider using the PSGBI in their work documenting the stability or change in environmental biases.
The development and validation of the PSGBI provides researchers, faculty mentors, and institutional interventionists with a new tool to assess and understand an important and difficult-to-study phenomenon within the complex academic setting. Our hope is that this project, and the PSGBI, continue to build on the existing conversation around creating less hostile work environments for women academics, especially within the STEM disciplines. We also hope that this project and index provides women academics and their allies with additional language to discuss women’s perceptions of subtle gender bias. And last, we hope that this index provides the women faculty, ally colleagues and administrators, interventionists, and other supporters with a set of events and experiences to understand where and when biases may be experienced, if not consciously perceived.
From the Center for Women and Work, we want to thank to Meg Sobcowicz-Kline, Marina Ruths, Michelle Haynes-Baratz, Laura Punnett, and Mignon Duffy for their support and work on this article and the larger project. From UMass Worcester Medical School, we want to especially thank Linda Churchill and Robert Milner for their dedication and support with the project and manuscript.
CS 340 Milestone One Guidelines and Rubric Overview: For this assignment, you will implement the fundamental operations of create, read, update,
Retail Transaction Programming Project Project Requirements: Develop a program to emulate a purchase transaction at a retail store. This
7COM1028 Secure Systems Programming Referral Coursework: Secure
Create a GUI program that:Accepts the following from a user:Item NameItem QuantityItem PriceAllows the user to create a file to store the sales receip
CS 340 Final Project Guidelines and Rubric Overview The final project will encompass developing a web service using a software stack and impleme
