Sage Journals: Discover world-class research

Abstract

Women’s underrepresentation in science, technology, engineering, and mathematics (STEM) is related to the hierarchical social structure of gender relations in these fields. However, interventions to increase women’s participation have focused primarily on women’s interests rather than on STEM managers’ hiring practices. In this research, we examine STEM hiring practices, explore the implicit bias in criteria used by STEM managers, and suggest possible corrective solutions. Using an experimental design with 213 men and women STEM managers, we show that when evaluating a female candidate, women and men STEM managers apply differential selection criteria, with men demonstrating implicit in-group gender favoritism in their hiring decisions. Specifically, the ability to work long hours was a more important criterion for male managers when evaluating a female candidate, forming an implicit gender bias, whereas female managers gave greater importance to problem-solving ability, a more gender-equal criterion. Adding a personal note to the curriculum vitae stating that the candidate had hired a full-time nanny was useful in decreasing the importance of the ability to work long hours criterion for men managers. We suggest individual and institutional interventions to reduce this bias, as a path to increasing women’s participation in STEM.

Keywords

hiring women STEM implicit gender bias in-group gender favoritism ability to long hours problem-solving ability

There is a major underrepresentation of women in the science, technology, engineering, and mathematics (STEM) fields) (Bird and Rhoton 2021). Men are employed in STEM fields twice as frequently as women (Landivar 2013). Gender disparities in transitions to STEM jobs are common in computer sciences and engineering even when women have degrees in these fields (Sassler et al. 2017). Although women’s participation in STEM has increased worldwide (Murphy and Oesch 2016), large gender disparities still remain (Mcquillan and Hernandez 2021). Increasing women’s representation in STEM promotes social justice and economic vibrancy (Kossek and Buzzanell 2018). Currently, the demand for STEM workers is increasing (Singh et al. 2018), and women can boost the pool of recruitment (Weisgram and Diekman 2015).

Gender inequality in STEM fields reflects prevalent social structures (Lockhart 2021), as science fields have been dominated by men for centuries, thus putting them in positions of power (Britton 2017). Organizational cultures and practices in STEM are socially constructed in ways that disadvantage women (Bird and Rhoton 2021; Cech 2013). Organizational practices limit women’s participation in STEM and originate from societal expectations that women have primary responsibility for child and family care and housekeeping. Thus, women are perceived as unable to work outside the home for as many hours as men (Kiser 2015). Furthermore, there is the tendency to underestimate women’s professional abilities along with an expectation that men are naturally inclined to science, mathematics, and practical pursuits—also challenging men who may not be inclined to gender-stereotypical roles (Eagly and Karau 2002; Hall et al. 2019).

Gender stereotypes are reflected in women’s self-bias toward their fit to STEM fields (Correll 2001) and in gender bias among men and women managers (Bird and Rhoton 2021). Most interventions have tried to increase women’s interest in science and confidence in their STEM abilities (Casad et al. 2018). However, fewer interventions focused on STEM managers, who make the hiring decisions (Botella et al. 2019). Yet STEM managers may discriminate against women in hiring decision making (Stamarski and Son Hing 2015). Diversity training aims to reduce implicit biases in managers’ perceptions of women’s and minority candidates’ abilities (Pritlove et al. 2019). However, this intervention was deemed ineffective in most cases (Dobbin and Kalev 2018).

The goals of this paper are, first, to identify an implicit bias in differential criteria of hiring a STEM project manager by men and women STEM managers, when evaluating two critical attributes of candidates: the ability to work long hours and problem-solving abilities. Second, we sought to test the effects of an intervention aimed at reducing the implicit ability to work long hours gender bias criterion among managers.

Figure 1 presents a conceptual model that summarizes the literature on the influence of explicit (conscious) and implicit (unconscious) perceptions and criteria on the hiring decision outcome, mainly focusing on implicit perceptions (e.g., women are unconsciously perceived to have lower ability to work long hours than men).

Figure 1:

Conceptual Model: The Sources of Gender Bias in the Hiring Process

To explore implicit criteria, we designed an experiment for men and women STEM managers in evaluating male and female curriculum vitae (CVs) (explicit perceptions) and the likelihood of hiring (outcome). Given that the ability to work long hours criterion was expected to be more important for men managers who are the dominant decision makers in STEM hires (Botella et al. 2019), we designed an intervention to reduce the importance of the ability to work long hours criterion. The treatment group received CVs with the addition of a personal note implying candidates’ ability to work long hours, whereas the control group did not. Using regression models, we found a differential use of the ability to work long hours criterion for men compared with women STEM managers in their evaluation of women candidates but not men candidates, and the intervention reduced the importance of the ability to work long hours criterion for men managers.

Our paper makes several contributions. First, we investigated an understudied implicit source of gender bias in hiring decisions of applicants, a criterion that favors male candidates. Second, ability to work long hours is an external, structural criterion that often disadvantages women, unrelated to the candidate’s skills. It is understudied in the context of hiring processes, although it is extensively studied in the “ideal worker literature” (e.g., Neely 2020; Padavic, Ely, and Reid 2020). We identify ability to work long hours as influencing hiring decisions and manipulated it by designing an intervention that reduced gender bias.

Third, the existing research shows contradictory findings regarding the hiring decisions of women managers toward women candidates. A body of literature suggests in-group gender favoritism, so that each gender favors candidates of their own gender in hiring decisions (e.g., Bendick and Nunes 2012). Other literature suggests that women managers also internalize the masculine point of view and implicitly favor men (e.g., Miller 2004). Our paper resolves this contradiction by showing that men and women managers highlight different hiring criteria for female candidates. Specifically, men emphasized a masculine criterion (ability to work long hours) while women emphasized a more gender-equal criterion (problem-solving ability).

Fourth, numerous interventions were designed to increase women’s participation in STEM. Most of these have targeted women, starting in their childhood—young girls, female students, and employees rather than managers (Botella et al. 2019). By targeting STEM managers, we aim to reveal the hidden organizational structures, cultures, and decision-making practices that are present in the STEM fields.

Theory and hypotheses

Historically, science fields possess androcentric values that are imbued with male biases (Thelwall, Abdullah, and Fairclough 2022). Women’s underrepresentation in STEM fields is still evident today (Bird and Rhoton 2021) and is related to the hierarchical nature of gender relations in these fields (Fox 2001), such as gender biases leading to discrimination in hiring and promotion (Wynn and Correll 2018). Because hiring is the initial stage in entering STEM, we focus on the hiring decisions of managers.

Sources of Gender Bias in Hiring Decisions Among Managers

Gender bias in the hiring process limits the proportion of women in male-dominated fields (Rice and Barth 2016). This bias derives from beliefs about gender-stereotypical social roles (Eagly and Wood 2012) that are defined within “culturally patterned relations” between a person and a social circle, influencing decisions and behaviors (Lopata and Thorne 1978). Thus, women are expected to care for children and to have stereotypical “feminine traits” (i.e., warmth, nurturing), and men are expected to fulfill instrumental roles at work. People tend to strive for congruence between stereotypes and job requirements, so women tend to be hired for female-dominated jobs consistent with gendered expectations (Eagly and Karau 2002).

Indeed, the outcome of hiring is that men are more frequently hired than women for prestigious high-income jobs (Koch, D’Mello, and Sackett 2015). This outcome may be influenced by explicit (conscious) or implicit (unconscious) biases (Blommaert, van Tubergen, and Coenders 2012). We define implicit bias as the general unconscious mental processes related to perceptions and decision-making processes that may produce discrimination (Greenwald and Krieger 2006) and affect organizational practices (Foley and Williamson 2019).

We built a conceptual model that divides hiring gender bias into two sources: biases rooted in the perceptions of a candidate’s abilities and biases rooted in the criteria affecting the hiring decision, which can be either explicit or implicit (see Figure 1). Bias in the perceptions of a candidate’s attributes is influenced by the worker’s gender rather than their performance (Isaac, Lee and Carnes 2009). Women explicitly received lower STEM attribute ratings than men even when their job qualifications are identical to or better than those of men (Reuben, Sapienza, and Zingales 2014).

But these perceptions can also be implicit when employers unconsciously perceive men to be more competent and better suited to science (Pritlove et al. 2019). This implicit bias is usually measured by the gender–science Implicit Association Test (IAT) and its variations (S. M. Jackson 2011; Nosek, Greenwald, and Banaji 2005). These tests measure the strength between implicit associations (e.g., women and science) and assume that individuals will respond more rapidly on a computer test when exposed to two associated concepts (Smeding 2012).

Similar to perceptions, we suggest that bias can be explicit or implicit in criteria that are different for the two genders (Rice and Barth 2017) or if both genders are evaluated based on a specific masculine criterion that men more commonly possess (Uhlmann and Cohen 2005). It is explicit when employers are aware of their criteria, or implicit when they are not aware (Greenwald and Krieger 2006).

Gender differences in implicit bias in perceptions have been extensively studied (e.g., Ebert, Steffens, and Kroth 2014), whereas gender differences in implicit bias in criteria have seldom been studied (Phelan, Moss-Racusin, and Rudman 2008). Asking managers what their criteria are for hiring men or women applicants may not necessarily yield the same results as implicitly assessing the criteria (Friedmann and Lowengart 2018). For example, managers can state explicitly that they use the same hiring criteria for women and men applicants, evaluate them similarly on ability to work long hours and problem-solving ability, and have the same hiring probabilities for both applicants. In this situation, there is no explicit gender bias. However, an implicit bias in hiring criteria may appear when men managers evaluate women candidates as high on ability to work long hours, and prefer to hire them rather than women they evaluate with low ability to work long hours. Such a finding may point to implicit bias used by managers to favor their gender.

Existing Interventions to Increase Women’s Hires in STEM Industries

Over the past two decades, the discourse on implicit bias and how to reduce it has grown (Nelson and Zippel 2021). For example, Wang and Degol (2017) maintain that it is not enough to focus on girls’ implicit bias toward math and science and women’s lack of personal fit to STEM careers (the supply of female workers for STEM). Interventions must focus on all relevant groups such as parents, teachers, and employers (who generate the demand for trained STEM workers). Indeed, most interventions focus on the initial stages of the “problem” when women developed their career aspirations and less on later stages focusing on the managers who make the hiring decisions (Botella et al. 2019).

The STEM organizational culture may also contribute to patterns of behaviors affecting the supply side (women’s choices) and the demand side (managers’ decisions). It is important to identify organizational mechanisms that hinder institutional changes (Cech 2013; Seron et al. 2018).

Dozens of interventions have been developed to increase women’s participation in STEM (e.g., Casad et al. 2018) focusing on the supply, demand, or culture. Such interventions include the following: (1) empowering women—interventions developed for girls and women in diverse settings, such as improving their skills, confidence, interest, and sense of belonging to the field (Friedmann 2018); (2) diversity training—workshops to reduce implicit gender bias that are directed at academic faculty, industry employees, and managers (Dobbin and Kalev 2016). These include delivering information about women’s barriers to improve attitudes toward women in STEM (S. M. Jackson, Hillard, and Schneider 2014). Diversity training was found questionably effective (Dobbin and Kalev 2018). Some researchers have even suggested that this training harms women’s interest in STEM because it emphasizes their minority status, exacerbating inequality (Wynn and Correll 2018). And (3) structural approaches include programs that foster pro-family policies or federal agencies that provide tax breaks for childcare expenses (Goulden, Mason, and Frasch 2011). These focus on organizational support to reduce gender bias in the long run (Madva 2020).

In a review, Nelson and Zippel (2021) applied the use of implicit bias as a concept to structural interventions in STEM academic frameworks. The implicit bias concept has five key features that promote its translatability into practice. It is demonstrable (there is empirical evidence), relatable (people have experienced this), versatile (can be relevant for different settings and organizations), actionable (applicable), and impartial (not assigning blame) (Nelson and Zippel 2021). These features are challenging for planning institutional change (such as hiring practices) if they do not reveal the hidden structural roots of gender bias. In this paper, we focus on implicit gender bias in STEM hiring practices, which aim to reveal some of the hidden sources of gender bias present in STEM organizational culture as reflected in managers’ hiring decisions.

Differences Between Men and Women Managers: Implicit Perceptions and Criteria

The literature on how men and women differ in gender biases is equivocal (Wynn and Correll 2018). Several experiments examined the implicit bias in perceptions using IAT, demonstrating that men perceive women to be less effective than men, but that women see themselves as equally competent as men (e.g., Ebert, Steffens, and Kroth 2014).

Some scholars have posited that women managers may adopt evaluation criteria used by their male counterparts (Miller 2004). Other scholars suggested that women and men behave differently as managers (Adams and Funk 2012) and that women managers are less gender biased (Rice and Barth 2017). Diversity training of managers has reduced men’s but not women’s implicit gender bias, suggesting that women may be less biased than men (Moss-Racusin et al. 2016). To investigate differences between women and men STEM managers in implicit gender bias in criteria, we built on Fishbein and Ajzen’s (1975) expectancy–value model, suggesting that hiring decisions are a function of explicit attribute perceptions and implicit criteria (i.e., attributes’ importance).

Explicit Perceptions and Implicit Criteria of Hiring Outcomes

In examining perceptions and criteria, we focus on (1) ability to work long hours, which is especially relevant in STEM (Kaushiva and Joshi 2020) and relates to organizational success (Van Dalen, Henkens, and Schippers 2010), and (2) problem-solving ability, which is the ability to find solutions to difficult or complex issues, specifically solving real-world STEM problems (Tati, Firman, and Riandi 2017). This ability is highly valued by STEM managers (Rayner and Papakonstantinou 2015). Very little is known about managers’ ability to work long hours and problem-solving ability perceptions and criteria. However, skills alone are not sufficient for work readiness (e.g. problem-solving ability), and candidates’ attributes and external work-related attributes (e.g. ability to work long hours) are both considered in the hiring process (D. Jackson 2016a).

The 24/7 STEM work ethic demands that the “ideal worker” can work long hours and is always available (Kaushiva and Joshi 2020). Men usually have an advantage in the ability to work long hours (W. M. Williams and Ceci 2012) because high-commitment careers are difficult to combine with family and housework responsibilities, particularly for women (Fritz and van Knippenberg 2018). Women, even those without children, often are perceived as having lower ability to work long hours than men because they are perceived as responsible for other domestic activities and duties (besides childcare), and may be viewed as future mothers (Bertogg et al. 2020).

Parents in general are perceived as less committed to their jobs, but parental status influences women more than men (Fuegen et al. 2004). In masculine-cultural organizations such as STEM, these stereotypes penalize mothers and may benefit fathers (Benard and Correll 2010). “Maternal-wall bias” is a form of organizational bias against mothers, including negative competence assumptions (J. C. Williams 2004). For example, pregnant women managers were perceived by employers as less committed to their jobs, and more irrational (Correll, Benard, and Paik 2007). The negative competence assumptions are higher for mothers of children younger than two years (Cuddy, Fiske, and Glick 2004). Women in their child-bearing years, with and without small children (Fuegen et al. 2004), encounter these stereotypes more often than older women and are the most penalized group in terms of wages (Correll, Benard, and Paik 2007). These negative perceptions discriminate against women in hiring, promotion, and salary decisions (Benard and Correll 2010). Thus, the working 24/7 STEM norm (Kaushiva and Joshi 2020), combined with the perception that women cannot work as long hours as men (Kelly et al. 2010), is expected to lead to implicit gender bias in criteria favoring men in STEM.

On the other hand, the criterion of problem-solving ability, which is central in any STEM field (Mitts 2016), implies a high level of function, where the solution is not obvious, and analytical thinking is required to solve the problem (Gomez and Albrecht 2014). Problem-solving abilities are comparable between both genders (Argaw et al. 2017; Else-Quest, Hyde, and Linn 2010). Problem-solving ability was one of the strongest hiring criteria among STEM employers (McGunagle and Zizka 2020).

Ability to work long hours benefits men structurally compared with women, and is something less controllable for women, whereas problem-solving ability is perceived as a gender-balanced ability. We examine the criteria and perceptions of both ability to work long hours and problem-solving ability attributes.

In-Group Favoritism

A social group is a collection of individuals with a shared identity (Turner et al. 1987). In-group favoritism, the preference for those who resemble “me” over out-group members, may influence hiring decisions (Fu et al. 2012; Tajfel and Turner 1979). Because the STEM fields tend to be masculine, a woman’s social identity as an outsider is salient (Richman, van Dellen, and Wood 2011). Thus, favoritism by men managers for men applicants might occur.

Many organizations are striving toward equity in their hiring decisions (Chesley 2011). Therefore, gender bias research has shifted from explicit to implicit (Uhlmann and Cohen 2005). We suggest that in-group gender favoritism has shifted to an implicit in-group gender favoritism by the use of a subtle strategy that fashions the hiring criteria to favor applicants of one gender in particular. Thus, we predict the following:

Hypothesis 1a: When evaluating the likelihood of hiring a woman candidate for a STEM job, the influence of ability to work long hours on hiring evaluations will be a more important criterion for men than for women managers (see Online Appendix A, Equation 1, $b_{4})$ .

Because the ability of women to work longer hours is perceived to be limited (Kiser 2015), we expect that women managers will evaluate women on a more balanced criterion related to job performance than would men managers—problem-solving ability. This would give women a fairer chance of entering STEM, unlike the ability to work long hours criterion that is disadvantageous for women (Luxton 2017). Thus, we predict the following:

Hypothesis 1b: When evaluating the likelihood of hiring a woman candidate for a STEM job, the influence of her problem-solving ability on hiring evaluation will be a more important criterion for women than for men managers (see Online Appendix A, Equation 1, $b_{5}) .$

Because whether the ability to work long hours is expected to be an important hiring criterion for men STEM managers, and because women candidates are perceived as having limited ability to work long hours (Kiser 2015), we focused our experimental intervention on ability to work long hours and not problem-solving ability. Therefore, we expect that the exposure of STEM men managers to a CV that promotes women’s ability to work long hours will reduce the importance of the ability to work long hours criterion. We predict a significant three-way interaction between gender of the manager, ability to work long hours rating, and the CV intervention note emphasizing ability to work long hours to demonstrate among men managers the lower importance of ability to work long hours criterion in the treatment group (with the note) compared with the control group (without the note).

Hence, we predict the following:

Hypothesis 2: When evaluating the likelihood of hiring a woman candidate for a STEM job, adding a note regarding ability to work long hours to the CV reduces the importance of the ability to work long hours criterion among men compared with women managers (see Online Appendix A, Equation 2, $b_{7})$

For the research model, see Figure 2.

Figure 2:

Research Model

Method

Pretest

Focus Group

Before the main study, we conducted an online focus group with five STEM managers (four men and one woman), which reflects the gender representation in STEM: 80% men, 20% women, M_age = 41 years, standard deviation (SD) = 1.58 (Committee on Equal Opportunities in Science and Engineering [CEOSE] 2015). The managers came from computer science and biorobotics industries from the center of Israel and were recruited through a snowball sampling. First, we asked them to identify the attributes important to them as managers considering a project manager hire. They suggested 11 attributes: the ability to solve STEM-related problems, work long hours, take initiative, apply technical knowledge, be time-efficient, learn new skills, work in a team, work independently, be assertive, present ideas in front of audiences, and take personal responsibility. The gender of the candidate was not mentioned as an important attribute. They chose ability to work long hours and problem-solving ability as the two most important attributes (similar to D. Jackson 2016b).

CV Intervention

We developed two versions of CVs, identical except for candidates’ gender, represented by gender-specific first names (see Online Appendix B). We chose names that implied majority-group race and class (Bertrand and Mullainathan 2004; Gaddis 2017). When designing the CV, we took five CVs that were submitted to a STEM manager’s position call and we used the same elements in the study’s fictional CVs. We created two versions because the use of an identical CV for different genders is not realistic, and we created two CVs in different STEM fields to generalize to STEM. The two versions were counterbalanced by fields (biopharma and biorobotics) and gender so that managers received two CVs to evaluate: a woman’s biopharma CV and a man’s biorobotics CV or vice versa. Each version of the CV contained equivalent academic background and was constructed mostly of instrumental achievements with some interpersonal management and experience. We did not note whether the candidate had children or not, because it is illegal to ask these questions in a job interview. These two versions were used as a control and treatment condition: control: CVs without any personal note; treatment: CVs with a personal note highlighting the ability to work long hours: “I am fully committed to my career, and therefore hired a full-time nanny to take care of family and household chores” (for both male and female CVs). See Table 1 for intervention manipulation check which confirmed that the note increased ability to work long hours perceptions but not the problem-solving ability perceptions.

Table 1:

Manipulation Check

	ControlM (SD)	TreatmentM (SD)	t-Test	p value
AWLH perception of a general candidate	3.62 (0.84)	4.03 (0.75)	–5.22	<.001
AWLH perception of a male candidate	3.58 (0.82)	4.02 (0.76)	–4.05	<.001
AWLH perception of a female candidate	3.66 (0.87)	4.03 (0.75)	–3.33	<.001
PSA perception of a general candidate	3.97 (0.84)	3.97 (0.74)	–0.54	=.96
PSA perception of a male candidate	4.01 (0.80)	3.96 (0.79)	0.45	=.65
PSA perception of a female candidate	3.92 (0.88)	3.98 (0.70)	–0.52	=.60

Note: Both candidates’ AWLH evaluations are higher with the note than without. AWLH = ability to work long hours; PSA = problem-solving ability.

Pretest of CVs

The STEM managers from the focus group were given the two versions, without the name and the gender of the applicant, and were asked to evaluate the candidate on the 11 attributes and hiring likelihood. No significant differences were found between the two versions, showing that both versions adequately represent STEM and were not biased in favor of one subfield.

Sample and Procedure

The sample size was determined using G*power for a multiple linear regression with 12 predictors. A minimum sample of 130 participants is required for power of 0.8, with 1-alpha = 95%, to detect a medium effect size (Cohen’s d = 0.15). We recruited 213 STEM managers from the United States and United Kingdom (38% women managers, M_age = 40.07, SD = 9.80) through Prolific¹ (treatment: n = 115, control: n = 98). Prolific is an online platform that recruits survey respondents of different demographic and work-related profiles for a modest fee. We prescreened for English language and managerial positions with hiring experience in STEM: computer sciences, engineering, chemistry, biochemistry, biomedical sciences, and physics.

This was a 2 (gender of the manager, between) by 2 (gender of the candidate, within) mixed research design. First, the STEM managers were randomly asked to read the CVs of two job applicants. After reading the CV, participants were asked three questions to ensure their attention to the CV’s details. Then they evaluated the candidates on their capacity for problem-solving ability and ability to work long hours, along with other attributes (to represent the evaluative process) and evaluated their hiring probabilities. Finally, respondents reported their demographics and were debriefed and paid.

Measures

Perception of ability to work long hours and problem-solving ability attributes were measured by asking “please rate the candidates’ ability to work long hours” and “please rate the candidates’ ability to solve problems at work.” Responses were rated on a 5-point Likert scale ranging from 1 (very low) to 5 (very high).

Outcome—Hiring evaluation probability was measured by asking “how likely is this candidate to be accepted for the position?” Responses were on a 5-point Likert scale ranging from 1 (not likely at all) to 5 (very likely). For demographics, we asked participants’ age, gender, family status, and education level.

Statistical Analyses

First, we computed correlations, means, and standard deviations for all variables and examined the effectiveness of the manipulation using a t-test. Then, to examine H1a and H1b, we performed linear regression. The dependent variable was the hiring evaluation of the male or female candidate. The independent variables were ability to work long hours, problem-solving ability, candidate gender × manager gender interaction while controlling for the specific field listed on the CV (biopharma or biorobotics), age, education, and family status, because these could influence their decisions (Gilbert and Malone 1995). The same analyses were performed for the treatment and control conditions.

Next, to examine H2, we used PROCESS, a modeling tool, to compute a set of regressions that fit the expected model (Hayes 2017). We ran a three-way interaction (model 3) between the gender of the manager × candidate’s ability to work long hours × CV intervention (G × ability to work long hours × CV) when predicting the hiring evaluation probabilities of either a female or male candidate, controlling for the same manager’s demographics as above. We used Hayes’ (2017) simple interaction (model 1) to illustrate the triple interaction results.

To examine whether the criteria are influenced by the candidates’ gender, we ran a mixed linear regression for men or women managers separately to examine the potential gender of the candidate × attribute (ability to work long hours or problem-solving ability) when predicting hiring decisions, for each condition of the CV intervention. This mixed regression analysis was done because each manager rated the job qualifications and hiring probabilities of a woman and man candidate in the same survey, so the random effect of participants’ identity was controlled for.

Finally, to test whether possible initial explicit gender bias in hiring outcome or ability to work long hours/problem-solving ability perceptions exists, we examined the control condition. We used a repeated-measures analysis of variance (ANOVA), where the manager’s gender served as a between-subject factor predicting the likelihood of hiring a female versus male candidate, so the hiring outcome evaluation of the two candidates served as the within-subject factor. The same analysis was used to examine explicit bias in the perceptions when the manager’s gender was the between-subject factor to predict either the ability to work long hours or problem-solving ability perceptions of female versus male candidates (within-subject factor).

Results

Correlations, means, and standard deviations are presented in Table 2. As shown in Table 3, in the control condition, when evaluating a female candidate, ability to work long hours was more important for men than for women managers (B _{manager’s gender _× ability to work long hours} = ‒0.38, SE = 0.18, p < .05). However, problem-solving ability was more important for women than for men managers (B _{manager’s gender _× problem-solving ability} = 0.45, SE = 0.20, p < .05). This confirmed H1a and H1b.

Table 2:

Correlations, Means, and Standard Deviations for the Control CV Intervention (CV without a Note) and Treatment CV Intervention (CV with a Note)

Correlations, means, and standard deviations for the control CV intervention (CV without a note).
Variable	Men managers											Women managers
Variable	1	2	3	4	5	6	7	8	9	M	SD	1	2	3	4	5	6	7	8	9	M	SD
1. STEM Field of the CV^a	—									1.48	0.50	—									1.51	0.50
2. Age of the manager	.00	—								28.92	6.47	–.02	—								32.58	7.91
3. Education of the manager^a	–.00	.25**	—							2.91	0.79	–.02	.48**	—							3.25	0.80
4. Family status of the manager^a	.02	.16*	.01	—						1.92	1.47	.00	.11	.13	—						2.5	1.5
5. Women’s candidate AWLH	–.17	.02	.16	.13	—					3.65	0.89	–.10	–.19	–.01	.12	—					3.69	0.82
6. Women’s candidate’s PSA	–.11	.13	.19	.15	.53**	—				3.86	0.93	–.12	–.28	–.11	.20	.58**	—				4.06	0.75
7. Male’s candidate AWLH	–.04	.13	.02	–.04	.55**	.37**	—			3.63	0.85	–.12	–.06	.10	–.11	.39**	.46**	—			3.47	0.74
8. Male’s candidate’s PSA	.06	–.01	.10	–.06	.40**	.47**	.48**	—		4.09	0.84	–.19	.07	–.09	.24	.20	.40*	.44**	—		3.83	0.70
9. Likelihood of hiring a woman	–.10	.22	.22	.19	.60**	.59**	.42**	.43**	—	3.70	0.90	.16	.13	.24	.28	.42**	.66*	.31	.50**	—	3.97	0.81
10. Likelihood of hiring a man	–.02	.08	.12	–.01	.55**	.37**	.55**	.59**	.58**	3.68	0.79	.26	–.03	–.00	.30	.33	.45**	.21	.45**	.02	3.81	0.62
Correlations, means, and standard deviations for the treatment CV intervention (CV with a note).
Variable	Men managers											Women managers
Variable	1	2	3	4	5	6	7	8	9	M	SD	1	2	3	4	5	6	7	8	9	M	SD
1. STEM Field of the CV^a	—									1.50	0.50	—									1.49	0.50
2. Age of the manager	.00	—								39.44	9.14	.01	—								40.78	10.52
3. Education of the manager^a	.00	.23*	—							3.02	0.82	–.01	.13	—							3.21	0.78
4. Family status of the manager^a	.00	.05	.12	—						2.52	1.43	.01	.20	–.14	—						3.00	1.63
5. Women’s candidate AWLH	–.03	–.02	.23*	.09	—					4.02	0.67	.03	.01	.30*	.02	—					4.02	0.84
6. Women’s candidate’s PSA	–.34*	.10	.16	–.20	.43**	—				4.04	0.69	.20	.17	.31*	.15	.39**	—				3.91	0.73
7. Male’s candidate AWLH	–.19	–.19	–.07	–.01	.25	.40**	—			4.02	–0.19	-.19	.18	.16	-.21	.52**	.39**				4.02	0.83
8. Male’s candidate’s PSA	.05	–.09	–.03	.02	.12	.32*	.30*	—		3.96	0.00	.10	-.06	.08	-.08	.52**	.48**	.45**			3.96	0.84
9. Likelihood of hiring a woman	–.33*	.19	.09	–.15	.28**	.62**	.34*	.38**	—	3.60	–0.19	-.19	.05	.12	.13	.44**	.63**	.65**	.52**		3.73	1.07
10. Likelihood of hiring a man	.37*	–.06	–.23	–.08	.03	.25	.16	.69**	.20	3.54	0.90	.00	–.06	.12	–.02	.49**	.40**	.66**	.69**	.64**	3.67	1.03

NOTE: STEM = science, technology, engineering, and mathematics; AWLH = ability to work long hours; PSA = problem-solving ability; CV = curriculum vitae.

STEM Field of the CV: 0 = Biopharma, 1 = Biorobotics; Education of the manager: 12 years, 13–15 years, 16–20 years, 20+ years; Family status of the manager 1 = single, 2 = married, 3 = divorced, 4 = widowed.

p < .05. **p < .01.

Table 3:

Regression Analysis When Predicting Hiring Evaluations of Women and Men Candidates: Control Condition (CV without a Note)

	Likelihood of hiringa woman			Likelihood of hiringa man
	Beta	SE	p value	Beta	SE	p value
Constant	–0.24	0.46	.61	0.38	0.48	.43
(c1) STEM field of the CV	0.07	0.13	.60	0.13	0.12	.26
(c2) Age of the manager	0.02**	0.01	.01	0.00	0.01	.75
(c3) Education of the manager	0.16	0.08	.05	0.11	0.08	.17
(c4) Family status of the manager	0.03	0.04	.48	–0.01	0.04	.85
(G) Manager’s gender ^a	–0.37	0.70	.59	1.12	0.74	.12
(PSA) Candidate’s PSA	0.29**	0.09	.00	0.35**	0.10	.00
(AWLH) Candidate’s AWLH	0.41**	0.09	.00	0.34**	0.09	.00
(G × PSA) Manager’s gender × PSA (H1b)	0.45*	0.20	.02	0.07	0.19	.71
(G × AWLH) Manager’s gender × AWLH (H1a)	–0.38*	0.18	.04	–0.34	0.18	.06
R²	.54**			.42**

NOTE: Values in bold represent tested hypotheses. STEM = science, technology, engineering, and mathematics; PSA = problem-solving ability; AWLH = ability to work long hours; CV = curriculum vitae.

p < .05. **p < .01.

Manager’s gender: 0 = men, 1 = women.

When evaluating a male candidate, the gender of the manager did not interact with problem-solving ability or the ability to work long hours. As can be seen in Table 4, in the treatment condition compared with the control, when the CVs included a note to deal with ability to work long hours, women candidates’ ability to work long hours was a significantly less important criterion for men managers than it was for women managers (B _{manager’s gender _× ability to work long hours} = 0.71, SE = 0.19, p < .001).

Table 4:

Regression Analysis When Predicting Hiring Evaluations of Women and Men Candidates: Treatment Condition (CV with a Note)

	Likelihood of hiringa woman			Likelihood of hiringa man
	Beta	SE	p value	Beta	SE	p value
Constant	1.28	0.81	.11	0.32	0.82	.70
(c1) STEM field of the CV	–0.21	0.13	.12	0.43	0.15	.06
(c2) Age of the manager	0.04	0.01	.51	–0.00	0.01	.63
(c3) Education of the manager	–0.19*	0.08	.02	–0.15	0.09	.11
(c4) Family status of the manager	0.01	0.04	.89	–0.01	0.04	.85
(G) Manager’s gender^a	–2.15*	0.87	.02	–0.63	0.90	.49
(PSA) Candidate’s PSA	0.83**	0.14	.00	0.72**	0.15	.00
(AWLH) Candidate’s AWLH	–0.81	0.14	.56	0.11	0.16	.50
(G × PSA) Manager’s gender × PSA	–0.11	0.20	.57	–0.07	0.20	.73
(G × AWLH) Manager’s gender × AWLH	0.71**	0.19	.00	0.27	0.22	.22
R²	.61**			.51**

NOTE: STEM = science, technology, engineering, and mathematics; PSA = problem-solving ability; AWLH = ability to work long hours; CV = curriculum vitae.

p < .05. **p < .01.

Manager’s gender: 0 = men, 1 = women.

The CV intervention succeeded in lowering the ability to work long hours (AWLH in equations) criterion for any candidate among men managers (B_{CV intervention} = ‒1.76, SE = 0.62, p < .001; B_AWLH = 0.32, SE = 0.12, p < .01; B_{CV intervention _× AWLH} = 0.35, SE = 0.16, p < .05), while it did not change the importance of the ability to work long hours criterion when evaluating any candidate among women managers (B _{CV intervention} = 0.47, SE = 0.52, p = .37; B_AWLH = 0.56, SE = 0.07, p < .001; B_{CV intervention _× AWLH} = ‒0.23, SE = 0.13, p = .08).

To confirm that the importance of the ability to work long hours criterion differs between the control and treatment conditions in the CV intervention, for the women and men managers, we examined the interaction of manager’s gender × candidate’s ability to work long hours × CV intervention. As shown in Table 5, this interaction was significant when evaluating hiring probabilities of a woman candidate (B _{manager’s gender _× AWLH _× CV intervention} = 0.79, SE = 0.28, p < .01), but not when evaluating a male candidate (B _{manager’s gender _× AWLH _× CV intervention} = 0.47, SE = 0.29, p = .13). See Figure 3 for the illustration of the manager’s gender × candidate’s ability to work long hours × CV intervention interaction. The intervention did not aim to change problem-solving ability (PSA in equations), so no change in the importance of this criterion was expected nor observed.

Table 5

Regression Analysis When Evaluating the Effect of the CV Intervention Among Women and Men Candidates Separately (Hayes Model 3)

	Likelihood of hiring a woman			Likelihood of hiring a man
	Beta	SE	p value	Beta	SE	p value
Constant	1.22	0.46	.09	1.28*	0.47	.01
(c1) STEM field of the CV	–1.95	0.11	.07	–0.003	0.01	.69
(c2) Age of the manager	0.02**	0.01	.02	0.02	0.07	.76
(c3) Education of the manager	0.07	0.07	.32	–0.03	0.03	.38
(c4) Family status of the manager	0.03	0.03	.40	1.42	0.74	.06
(G) Manager’s gender^a	0.78	0.69	.26	1.12	0.74	.12
(AWLH) Candidate’s AWLH	0.56**	0.10	.00	0.53**	0.10	.00
(CV) CV intervention^b	1.05	0.74	.16	0.04	0.75	.95
(G x AWLH) Manager’s gender × AWLH	–0.16	0.18	.40	–0.32	0.20	.11
(G × CV) Manager’s gender × CV intervention	–3.28**	1.10	.00	–1.18	1.13	.12
(CV × AWLH) CV intervention × candidate’s AWLH	–0.37*	0.19	.05	–0.07	0.19	.72
(G × AWLH × CV) Manager’s gender × candidate’s AWLH × CV intervention (H2)	0.79**	0.28	.005	0.47	0.29	.13
R²	.34**			.24**

Note: Values in bold represent tested hypotheses. STEM = science, technology, engineering, and mathematics; AWLH = ability to work long hours; CV = curriculum vitae.

p < .05. **p < .01.

Manager’s gender: 0 = men, 1 = women.

CV intervention: 0 = control, 1 = treatment.

Figure 3:

The Moderating Role of CV Intervention and Manager’s Gender on AWLH Importance When Predicting the Hiring Evaluation of a Female Candidate

As Table 6 shows, when predicting the hiring evaluation of a female candidate for men managers, the intervention decreased the importance of ability to work long hours criterion, while unexpectantly the intervention increased the importance of ability to work long hours criterion for women managers. As expected, the male candidate’s evaluation did not yield a significant interaction effect.

Table 6:

Regression Analysis for Managers, Evaluating Candidates, Separately per Gender (Hayes Model 1)

Men managers	Likelihood of hiring a woman			Likelihood of hiring a man
	Beta	SE	p value	Beta	SE	p value
Constant	0.38	0.49	.44	1.19	0.51	.02
(c1) STEM field of the CV	0.26	0.14	.06	0.35	0.13	.01
(c2) Age of the manager	0.02	0.01	.02	0.00	0.01	.75
(c3) Education of the manager	0.09	0.08	.29	–0.01	0.08	.87
(c4) Family status of the manager	–0.01	0.04	.80	–0.04	0.05	.37
(AWLH) Candidate’s AWLH	0.56	0.09	.00	0.53	0.10	.00
(CV) CV intervention	0.98	.073	.18	–0.04	0.74	.96
(CV × AWLH) CV intervention × candidate’s AWLH	–0.36	0.18	.04	–0.06	0.18	.73
R²	.31**			.27**
Women managers	Likelihood of hiring a woman			Likelihood of hiring a man
	Beta	SE	p value	Beta	SE	p value
Constant	1.84	0.79	.02	2.89	0.85	.00
(c1) STEM field of the CV	0.08	0.18	.67	0.29	0.19	.12
(c2) Age of the manager	0.01	0.01	.43	–0.01	0.01	.38
(c3) Education of the manager	0.03	0.12	.80	0.05	0.12	.68
(c4) Family status of the manager	0.09	0.06	.11	–0.01	0.06	.88
(AWLH) Candidate’s AWLH	0.39	0.16	.02	0.18	0.19	.34
(CV) CV intervention^a	–2.28	0.86	.06	–1.76	0.92	.07
(CV × AWLH) CV intervention × candidate’s AWLH	0.45	0.22	.04	0.39	0.24	.11
R²	.40**			.19*

Note: STEM = science, technology, engineering, and mathematics; AWLH = ability to work long hours; CV = curriculum vitae.

p < .05. **p < .01.

CV intervention: 0 = control, 1 = treatment.

The mixed regression analysis was performed on women and men managers separately and examined whether they used different criteria based on the candidate’s gender. The women and men managers assigned equivalent levels of importance to the attributes of the woman and man job candidates in the control condition (Male managers: B_{Candidate’s gender _× AWLH} = 0.06, SE = 0.14, p = .63; M_{Candidate’s gender _× PSA} = ‒0.04, SE = 0.14, p = .79; Women managers: B_{Candidate’s gender _× AWLH} = 0.005, SE = 0.20, p = .98; M_{Candidate’s gender _× PSA} = 0.33, SE = 0.21, p = .13) and in the treatment condition (men managers: B_{Candidate’s gender _× AWLH} = ‒0.21, SE = 0.20, p = .30; M _{Candidate’s gender _× PSA} = 0.20, SE = 0.20, p = .31; women managers: B_{Candidate’s gender _× AWLH} = 0.32, SE = 0.21, p = .14; M_{Candidate gender’s _× PSA} = 0.04, SE = 0.23, p = .86). This means the criteria were stable across candidates.

The CV intervention succeeded in lowering the ability to work long hours criterion for any candidate among men (B_{CV intervention} = ‒1.76, SE = 0.62, p < .001; B_AWLH = 0.32, SE = 0.12, p < .01; B_{CV intervention _× AWLH} = 0.35, SE = 0.16, p < .05), while it did not change the importance of the ability to work long hours criterion when evaluating any candidate among women managers (B_{CV intervention} = 0.47, SE = 0.52, p = .37; B_AWLH = 0.56, SE = 0.07, p < .001; B_{CV intervention _× AWLH} = ‒0.23, SE = 0.13, p = .008). See Figure 4 for the summary of results for the ability to work long hours criterion.

Figure 4:

The Importance of AWLH Criterion, with or without a Personal Note for Men and Women Managers, Evaluating Candidates

To test whether possible initial gender bias in explicit ability to work long hours perceptions was evident, we found that the interaction was insignificant, F(1, 107) = 1.03, p = .31. Managers did not perceive women and men candidates differently on the ability to work long hours attribute (men managers: M_{Women’s AWLH} = 3.65, SD = 0.90; M_{Men’s AWLH} = 3.68, SD = 0.80; women managers: M_{Women’s AWLH} = 3.69, SD = 0.82; M_{Men’s AWLH} = 3.47, SD = 0.73). The same pattern was evident in the problem-solving ability perceptions, and the interaction was insignificant, F(1, 107) = 2.37, p = .13. Managers did not perceive women and men candidates differently on the problem-solving ability attribute (men managers: M_{Women’s PSA} = 3.87, SD = 0.94; M_{Men’s PSA} = 4.10, SD = 0.77; women managers: M_{Women’s PSA} = 4.06, SD = 0.75; M_{Men’s PSA} = 3.81, SD = 0.71). Thus, we did not find any explicit gender bias in perceptions hiring.

Last, to test whether possible initial gender bias in the hiring decision outcome exists, we found an insignificant interaction between the factor (hiring male vs. female candidate) by the manager’s gender: F(1, 107) = 0.009, p = .92 (men: M_{Hiring women} = 3.69, SD = 0.89; M_{Hiring men} = 3.70, SD = 0.76; women: M_{Hiring women} = 3.97, SD = 0.81; M_{Hiring men} = 3.72, SD = 0.66). These results indicate that in general, men were not preferred over women, and that there was no explicit bias in the initial hiring outcome.

Discussion

Our goal was to explore implicit and explicit bias in hiring practices in STEM focusing on hiring criteria, specifically ability to work long hours and problem-solving ability among men and women managers. We tested an intervention to deal with one criterion (ability to work long hours) that particularly affects women as they seek entry into STEM work.

We exposed an implicit gender bias in a specific criterion, as men managers gave greater importance than women managers to the ability to work long hours, and displayed a type of implicit in-group gender favoritism. This was not explicit discrimination because the importance of the ability to work long hours criterion was exposed indirectly when examining explicit ability to work long hours and problem-solving ability perceptions on the hiring outcome, while there were no significant differences between men and women candidates in the explicit ability to work long hours and problem-solving ability perceptions. These findings stress the hidden androcentric values in STEM (Thelwall, Abdullah, and Fairclough 2022) and reveal the salience of gender biases in STEM hiring practices.

Woman managers (compared with men managers) gave greater importance to problem-solving ability, which is a more gender-balanced criterion, as problem-solving ability was conceptualized as concrete STEM-related problem-solving. This criterion allowed women with relevant skills to be considered competent for the job. By contrast, women managers attributed less weight to the ability to work long hours criterion, which challenges many women.

The ability to work long hours intervention reduced men managers’ importance of the ability to work long hours criterion; however, it elevated women managers’ importance of the ability to work long hours, as they expected the female candidate to have ability to work long hours capacity. This is a type of gender backlash: Women who possess masculine attributes are more severely judged due to their inconsistency with gender expectations (Rudman and Fairchild 2004). This “win some, lose some” consequence was unexpected, suggesting that women managers may unconsciously self-correct their hiring criteria according to the STEM (male-oriented) norms. This consequence should be kept in mind, as women managers are still the minority in STEM.

Our research is supported by the claim that the 24/7 work criterion is more achievable for men than for women, and this sustains workplace inequality (Padavic, Ely, and Reid 2020). Thus, the bias we found in the form of differential criteria for women and men STEM managers, referring to female candidates, is a gender bias that perpetuates social power structures (Nelson and Zippel 2021). However, one must wonder about how this criterion became implicitly important for men managers today. There are two possible explanations: (1) Women’s perceived low ability to work long hours caused this structural disadvantage, which in turn leads to discrimination in the hiring process, or (2) discrimination against women is a priori, and justified by their perceived low ability to work long hours. In this case, if we address this criterion individually to reduce hiring disparities, then we might bring about the selection of a different criterion that will justify the discrimination. The two explanations together create a circular reality, in which it is not clear how this implicit ability to work long hours normative criterion was institutionalized for men managers, but it is unlikely to change unless structural mechanisms address the needs of all workers.

Theoretical Contributions

Our research has several theoretical contributions. First, we show that an implicit gender bias in criteria is present in STEM hiring. Past literature has suggested explicit and implicit gender biases, focusing on implicit gender bias in perceptions of candidates’ abilities using the IAT test. We differentiated between two forms of implicit biases used for hiring decisions when referring to implicit bias in perceptions and criteria. Second, this paper is the first to demonstrate the importance of an external attribute: ability to work long hours, in a theoretical model that was empirically tested. Third, existing literature was equivocal regarding whether women and men STEM managers both internalize the masculine point of view in the criteria they use for hiring decisions. We show that they hold different criteria, which fits the prediction of in-group favoritism model in a subtler form. Thus, we suggest the implicit in-group gender favoritism model. Only the men managers kept the ability to work long hours barrier for hiring female candidates, whereas the women managers focused on problem-solving ability, which does not give one gender an advantage. Fourth, we targeted managers (instead of candidates) and demonstrated a successful intervention to deal with gender bias.

Practical Implications

This paper makes four practical contributions, two individual and two organizational. First, women can take the perspective of managers and choose to specify their ability to work long hours in the hiring processes, in the form of a personal note or a comment made in an interview. Neumark (2018) recommended that women should add job-relevant information to their CV because their gender may influence the manager’s hiring decision. However, this recommendation highlights the women’s ability to work long hours paradox: While the recommended note may ultimately reduce implicit gender bias in STEM hires, it may explicitly highlight inequality and preserve traditional gender expectations, by acknowledging that women do not usually have the ability to work long hours. This is aligned with Judith Lorber’s Paradoxes of Gender: women’s fight to erase the effects of sex differences, invokes them (Lorber 1994). The recommendation to add this note assumes gender inequality, as found in most STEM organizations (Thébaud and Charles 2018). Thus, in the STEM fields, the note is expected to be beneficial, but is not without hidden costs.

Second, managers should be aware of ability to work long hours as an implicit bias criterion in making hiring decisions, because sometimes awareness of the implicit bias (e.g., through professional journals or in diversity training) can bring about individual attitudinal change (Forscher et al. 2019). If they design the training, they should provide information on this type of implicit bias. This will make the implicit bias more demonstrable and relatable (Nelson and Zippel 2021).

Individual changes among candidates and managers may be effective in the short run, but this might individualize inequality while leaving structural roots of gender bias hidden (Nelson and Zippel 2021). Thus, we offer two additional organizational suggestions that strive to deal with the roots of the problem. The first involves accepting the need of the organization for employees who can work long hours and supporting employees by adding salary to cover home and child care expenses for all employees, similar to travel expenses. Intel introduced an intervention during COVID-19 lockdowns that provided all employees with activities for entertaining children and reimbursement for family care. This reduced work–family imbalance related to COVID-19 (Chung et al. 2021). Such structural interventions emphasize that women are not solely responsible for home and childcare chores, thus dealing with the roots of the social problem.

The other organizational change is to challenge the need for ability to work long hours by decreasing the demand for over-long work hours. This expectation of long hours may be how gender discrimination is embedded in the organization itself, illustrating how inequalities between women and men continue despite attempts to erase them (Acker 2006). This change of norms in STEM fields might dismantle a hidden structural barrier for women.

Conclusion

Although the note in the CV reduced implicit bias for men managers, there might be other types of biases that were not measured here. For example, we examined two attributes in hiring decisions, but other attributes, such time efficiency or interpersonal skills, might also be related to gender bias. We also manipulated only ability to work long hours, whereas future interventions might manipulate other attributes (such as problem-solving ability). It is possible that women’s adding a personal note highlighting exceptional problem-solving ability may increase the importance of this criterion as compared with ability to work long hours. We do not claim that this is the sole source of implicit gender bias in STEM, because women’s interest in the field, educational choices, or prior relevant work experience may also handicap their hiring chances. Future research should also focus on women’s implicit bias toward their fit to STEM careers.

The use of a nanny note in the CV relates to parents seeking STEM employment, and interventions of this type should include household help, which is applicable to both men and women hires. Both men and women have additional needs (e.g., housekeeping, care of pets, care of elderly). A “domestic help” note would increase the generalizability of the findings. Examining the effect of other types of domestic help can also clarify whether ability to work long hours is more about motherhood/fatherhood than about gender. If not, future studies should explore the impact of a note stating that the partner has job flexibility allowing him responsibility for domestic chores.

Other issues to explore include how the importance of the ability to work long hours criterion may change for older/younger candidates who are parents of young or older children. It might be that the age of the candidate or age of children will be more relevant than actual parenting status, as young candidates can potentially have children in the future. Future studies should explore the actual effectiveness of this intervention using real hiring situations, beyond the use of experimental design that has disadvantage of limited external validity.

In our research, the note was written based on the assumption that gender is viewed as binary. However, nowadays, gender is conceptualized on a broader continuum, including people with undefined gender (Alfrey and Twine 2017). Future studies should consider recommendations for the employment of these populations beyond the traditional gender dichotomy. In addition, many STEM hiring decisions depend on hiring committees, not just on individual hiring decisions. Thus, the group decision context may be further studied.

Some women in the STEM fields might not wish to prioritize career over family or utilize the salary budget for childcare. Future studies should explore the proportion of women who feel comfortable adding this note or outsourcing home and childcare chores. The intervention suggested might be class biased: Only those who can afford a nanny can add such a note. However, the high salary in STEM fields might suggest this recommendation is less class relevant (Bannikova and Petrov 2014).

This study did not control for many managers’ characteristics such as level of religiosity, race, and religion, since most STEM professionals are relatively homogeneous in their class, mainly Caucasian, high socioeconomic status, men (Ma and Liu 2015; McMorris 2016). Future studies should further control for less stereotyped STEM managers because the results may vary for them.

The experiment was conducted before COVID-19, which has turned into a real-life intervention that might lower the importance of AWLH as employees increasingly worked from home. Future research should explore whether the COVID-19 pandemic has altered hiring criteria in STEM relating to long hours or attendance in person in the workplace.

In summary, the low percentage of women working in STEM careers reflects implicit gender bias (Reuben, Sapienza, and Zingales 2014). We found that when evaluating female candidates, men managers have an implicit in-group gender favoritism. We further suggest that individual and institutional interventions may reduce the importance managers assign to the ability to work long hours criterion. This paper adds to the previous literature, which recommends interventions for women rather than changing the implicit gendered culture in STEM hires.

Supplemental Material

sj-docx-1-gas-10.1177_08912432221137910 – Supplemental material for Gender Bias in Stem Hiring: Implicit In-Group Gender Favoritism Among Men Managers

Supplemental material, sj-docx-1-gas-10.1177_08912432221137910 for Gender Bias in Stem Hiring: Implicit In-Group Gender Favoritism Among Men Managers by Enav Friedmann and Dorit Efrat-Treister in Gender & Society

Footnotes

Authors’ note:

We thank Gal Gutman, Niza Yanay, and Idit Fast for reviewing the paper, and giving their insightful comments that helped us improve our work.

ORCID iD

Enav Friedmann

Supplemental Material

Supplemental material for this article is available online.

Notes

Enav Friedmann, Ph.D. Assistant professor at the Guilford Glazer Faculty of Business and Management at Ben Gurion University (BGU) in Israel. Enav is the head of the BGU marketing lab. She holds a Ph.D. in Business Administration from BGU and was a visiting scholar at Ca’ Foscari University in Venice, Italy. Her current research includes brand preferences and purchasing choices, tailoring to heterogeneous consumer strata, specifically gender-related marketing, and social marketing.

Dorit Efrat-Treister is a Senior Lecturer at the Guilford Glazer Faculty of Business and Management, Ben-Gurion University of the Negev. She received her Ph.D. in Organizational Behavior at the Technion, Israel Institute of Technology, and continued as a postdoctoral fellow at the Sauder School of Business, University of British Columbia.

References

Acker

Joan

. 2006. Inequality regimes: Gender, class, and race in organizations. Gender & Society 20 (4): 441–64.

Adams

Renée B.

Funk

Patricia

. 2012. Beyond the glass ceiling: Does gender matter? Management Science 58 (2): 219–35.

Alfrey

Lauren

Twine

France Winddance

. 2017. Gender-fluid geek girls: Negotiating inequality regimes in the tech industry. Gender & Society 31 (1): 28–50.

Argaw

Aweke Shishigu

PetHailerov

Beyene Bashu

Ayalew

Beyene Tesfaw

Kuma

Shiferaw Gadisa

2017. The effect of problem based learning (PBL) instruction on students’ motivation and problem solving skills of physics. Eurasia Journal of Mathematics, Science and Technology Education 13 (3): 857–71.

Bannikova

Liudmila

Petrov

Alexander

. 2014. Women in engineering careers. Paper presented at 7th International Conference of Education, Research and Innovation (ICERI), Seville, Spain.

Benard

Stephen

Correll

Shelley J.

2010. Normative discrimination and the motherhood penalty. Gender & Society 24 (5): 616–46.

Bendick

Marc

Jr. Nunes

Anna P.

2012. Developing the research basis for controlling bias in hiring. Journal of Social Issues 68 (2): 238–62.

Bertogg

Ariane

Imdorf

Christian

Hyggen

Christer

Parsanoglou

Dimitris

Stoilova

Rumiana

. 2020. Gender discrimination in the hiring of skilled professionals in two male-dominated occupational fields: A factorial survey experiment with real-world vacancies and recruiters in four European countries. KZfSS Kölner Zeitschrift für Soziologie und Sozialpsychologie 72 (1): 261–89.

Bertrand

Marianne

Mullainathan

Sendhil

. 2004. Are Emily and Greg more employable than Lakisha and Jamal? A field experiment on labor market discrimination. American Economic Review 94 (4): 991–1013.

10.

Bird

Sharon R.

Rhoton

Laura A.

2021. Seeing isn’t always believing: Gender, academic STEM, and women scientists’ perceptions of career opportunities. Gender & Society 35 (3): 422–48.

11.

Blommaert

Lieselotte

van Tubergen

Frank

Coenders

Marcel

. 2012. Implicit and explicit interethnic attitudes and ethnic discrimination in hiring. Social Science Research 41 (1): 61–73.

12.

Botella

Carmen

Rueda

Silvia

López-Iñesta

Emilia

Marzal

Paula

. 2019. Gender diversity in STEM disciplines: A multiple factor problem. Entropy 21 (1): 30.

13.

Britton

Dana M.

2017. Beyond the chilly climate: The salience of gender in women’s academic careers. Gender & Society 31 (1): 5–27.

14.

Casad

Bettina J.

Oyler

Danielle L.

Sullivan

Erin T.

McClellan

Erika M.

Tierney

Destiny N.

Anderson

Drake A.

Greeley

Paul A

Fague

Michael A.

Flammang

Brian J.

2018. Wise psychological interventions to improve gender and racial equality in STEM. Group Processes & Intergroup Relations 21 (5): 767–87.

15.

Cech

Erin A.

2013. The self-expressive edge of occupational sex segregation. American Journal of Sociology 119 (3): 747–89.

16.

Committee on Equal Opportunities in Science and Engineering. 2015. Broadening participation in STEM: 2013–2014 biennial report to congress. https://beta.nsf.gov/funding/initiatives/broadening-participation/reports.

17.

Chesley

Noelle

. 2011. Stay-at-home fathers and breadwinning mothers: Gender, couple dynamics, and social change. Gender & Society 25 (5): 642–64.

18.

Chung

Heejung

Birkett

Holly

Forbes

Sarah

Seo

Hyojin

. 2021. COVID-19, flexible working, and implications for gender equality in the United Kingdom. Gender & Society 35 (2): 218–32.

19.

Correll

Shelley J.

2001. Gender and the career choice process: The role of biased self-assessments. American Journal of Sociology 106 (6): 1691–1730.

20.

Correll

Shelley J.

Benard

Stephen

Paik

. 2007. Getting a job: Is there a motherhood penalty? American Journal of Sociology 112 (5): 1297–1338.

21.

Cuddy

Amy J. C.

Fiske

Susan T.

Glick

Peter

. 2004. When professionals become mothers, warmth doesn’t cut the ice. Journal of Social Issues 60 (4): 701–18.

22.

Dobbin

Frank

Kalev

Alexandra

. 2016. Why diversity programs fail. Harvard Business Review 94 (7): 14. https://hbr.org/2016/07/why-diversity-programs-fail.

23.

Dobbin

Frank

Kalev

Alexandra

. 2018. Why doesn’t diversity training work? The challenge for industry and academia. Anthropology Now 10 (2): 48–55.

24.

Eagly

Alice H.

Karau

Steven J.

2002. Role congruity theory of prejudice toward female leaders. Psychological Review 109 (3): 573–98.

25.

Eagly

Alice H.

Wood

Wendy

. 2012. Social role theory. In Handbook of theories in social psychology, edited by Van Lange

Paul A. M.

Kruglanski

Arie W.

Tory Higgins

London: SAGE.

26.

Ebert

Irena D.

Steffens

Melanie C.

Kroth

Alexandra

. 2014. Warm, but maybe not so competent?—Contemporary implicit stereotypes of women and men in Germany. Sex Roles 70 (9–10): 359–75.

27.

Else-Quest

Nicole M.

Hyde

Janet Shibley

Linn

Marcia C.

2010. Cross-national patterns of gender differences in mathematics: A meta-analysis. Psychological Bulletin 136 (1): 103–27.

28.

Fishbein

Martin

Ajzen

Icek

. 1975. Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley.

29.

Foley

Meraiah

Williamson

Sue

. 2019. Managerial perspectives on implicit bias, affirmative action, and merit. Public Administration Review 79 (1): 35–45.

30.

Forscher

Patrick S.

Lai

Calvin K.

Axt

Jordan R.

Ebersole

Charles R.

Herman

Michelle

Devine

Patricia G.

Nosek

Brian A.

2019. A meta-analysis of procedures to change implicit measures. Journal of Personality and Social Psychology 117 (3): 522–59.

31.

Fox

Mary Frank

. 2001. Women, science, and academia: Graduate education and careers. Gender & Society 15 (5): 654–66.

32.

Friedmann

Enav

. 2018. Increasing women’s participation in the STEM industry. Journal of Social Marketing 8 (4): 442–60.

33.

Friedmann

Enav

Lowengart

Oded

. 2018. The context of choice as boundary condition for gender differences in brand choice considerations. European Journal of Marketing 52 (5/6): 1280–1304.

34.

Fritz

Claudia

van Knippenberg

Daan

. 2018. Gender and leadership aspiration: The impact of work–life initiatives. Human Resource Management 57 (4): 855–68.

35.

Feng

Tarnita

Corina E.

Christakis

Nicholas A.

Wang

Long

Rand

David G.

Nowak

Martin A.

2012. Evolution of in-group favoritism. Scientific Reports 2:460–65.

36.

Fuegen

Kathleen

Biernat

Monica

Haines

Elizabeth

Deaux

Kay

. 2004. Mothers and fathers in the workplace: How gender and parental status influence judgments of job-related competence. Journal of Social Issues 60 (4): 737–54.

37.

Gaddis

S. Michael

. 2017. How black are Lakisha and Jamal? Racial perceptions from names used in correspondence audit studies. Sociological Science 4:469–89.

38.

Gilbert

Daniel T.

Malone

Patrick S.

1995. The correspondence bias. Psychological Bulletin 117 (1): 21–38.

39.

Gomez

Alan

Albrecht

Bryan

. 2014. True STEM education. Technology and Engineering Teacher 73 (4): 8–16.

40.

Goulden

Marc

Mason

Mary Ann

Frasch

Karie

. 2011. Keeping women in the science pipeline. Annals of the American Academy of Political and Social Science 638 (1): 141–62.

41.

Greenwald

Anthony G.

Krieger

Linda Hamilton

. 2006. Implicit bias: Scientific foundations. California Law Review 94 (4): 945–67.

42.

Hall

Erika V.

Hall

Alison V.

Galinsky

Adam D.

Phillips

Katherine W.

2019. MOSAIC: A model of stereotyping through associated and intersectional categories. Academy of Management Review 44 (3): 643–72.

43.

Hayes

Andrew F.

2017. Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. 2nd ed. New York: Guilford Press.

44.

Isaac

Carol

Lee

Barbara

Carnes

Molly

. 2009. Interventions that affect gender-bias in hiring: A systematic review. Academic Medicine: Journal of the Association of American Medical Colleges 84 (10): 1440–46.

45.

Jackson

Denise

. 2016a. Modelling graduate skill transfer from university to the workplace. Journal of Education and Work 29 (2): 199–231.

46.

Jackson

Denise

. 2016b. Re-conceptualizing graduate employability: The importance of pre-professional identity. Higher Education Research & Development 35 (5): 925–39.

47.

Jackson

Sarah Marie.

2011. Assessment of implicit attitudes toward women faculty in science, technology, engineering, and math. Ph.D. diss., Wright State University, Dayton, OH.

48.

Jackson

Sarah Marie

Hillard

Amy L.

Schneider

Tamera R.

2014. Using implicit bias training to improve attitudes toward women in STEM. Social Psychology of Education 17 (3): 419–38.

49.

Kaushiva

Pavni

Joshi

Chetan

. 2020. Women’s re-entry into workforce: Experiences from India. In The new ideal worker: Organizations between work-life balance, gender and leadership, edited by Maestro

Mireia las Heras

Albiol

Nuria Chinchilla

Grau

Marc Grau

. Cham, Switzerland: Springer.

50.

Kelly

Erin L.

Ammons

Samantha K.

Chermack

Kelly

Moen

Phyllis

. 2010. Gendered challenge, gendered response: Confronting the ideal worker norm in a white-collar organization. Gender & Society 24 (3): 281–303.

51.

Kiser

Angelina I. T.

2015. Workplace and leadership perceptions between men and women. Gender in Management: An International Journal 30 (8): 598–612.

52.

Koch

Amanda J.

D’Mello

Susan D.

Sackett

Paul R.

2015. A meta-analysis of gender stereotypes and bias in experimental simulations of employment decision making. Journal of Applied Psychology 100 (1): 128–61.

53.

Kossek

Ellen Ernst

Buzzanell

Patrice M.

2018. Women’s career equality and leadership in organizations: Creating an evidence-based positive change. Human Resource Management 57 (4): 813–22.

54.

Landivar

Liana Christin

. 2013. Disparities in STEM Employment by sex, race, and Hispanic origin. Washington, DC: U.S. Census Bureau. https://www.census.gov/library/publications/2013/acs/acs-24.html

55.

Lockhart

Jeffrey W.

2021. Paradigms of sex research and women in STEM. Gender & Society 35 (3): 449–75.

56.

Lopata

Helena Z.

Thorne

Barrie

. 1978. On the term “sex roles.” Signs: Journal of Women in Culture and Society 3 (3): 718–21.

57.

Lorber

Judith

. 1994. Paradoxes of Gender. New Haven, CT: Yale University Press.

58.

Luxton

Meg

. 2017. Valuing unpaid work in the home. Policy briefing note. Canadian Research Institute for the Advancement of Women. https://www.criaw-icref.ca/wp-content/uploads/2021/04/Valuing-Unpaid-Work-in-the-Home.pdf.

59.

Yingyi

Liu

Yan

. 2015. Race and STEM degree attainment. Sociology Compass 9 (7): 609–18.

60.

Madva

Alex

. 2020. Individual and structural interventions. In An introduction to implicit bias: Knowledge, justice, and the social mind, edited by Beeghly

Erin

Madva

Alex

. New York: Routledge.

61.

McGunagle

Doreen

Zizka

Laura

. 2020. Employability skills for 21st-century STEM students: The employers’ perspective. Higher Education, Skills and Work-based Learning 10 (3): 591–606.

62.

McMorris

Jennifer Elise.

2016. The role of religion and gender in shaping STEM education and workforce participation. Ph.D. diss., The University of Texas at Austin, Austin, TX.

63.

Mcquillan

Julia

Hernandez

Nestor

. 2021. Real-life conundrums in the struggle for institutional transformation. Gender & Society 35 (3): 300–29.

64.

Miller

Gloria E.

2004. Frontier masculinity in the oil industry: The experience of women engineers. Gender, Work & Organization 11 (1): 47–73.

65.

Mitts

Charles R.

2016. Why STEM? Technology and Engineering Teacher 75 (6): 30–35.

66.

Moss-Racusin

Corinne A.

van der Toorn

Jojanneke

Dovidio

John F.

Brescoll

Victoria L.

Graham

Mark J.

Handelsman

. 2016. A “scientific diversity” intervention to reduce gender-bias in a sample of life scientists. CBE Life Sciences Education 15 (3): 1–11.

67.

Murphy

Emily

Oesch

Daniel

. 2016. The feminization of occupations and change in wages: A panel analysis of Britain, Germany, and Switzerland. Social Forces 94 (3): 1221–55.

68.

Neely

Megan Tobias

. 2020. The portfolio ideal worker: Insecurity and inequality in the new economy. Qualitative Sociology 43 (2): 271–96.

69.

Nelson

Laura K.

Zippel

Kathrin

. 2021. From theory to practice and back: How the concept of implicit bias was implemented in academe, and what this means for gender theories of organizational change. Gender & Society 35 (3): 330–57.

70.

Neumark

David

. 2018. Experimental research on labor market discrimination. Journal of Economic Literature 56 (3): 799–866.

71.

Nosek

Brian A.

Greenwald

Anthony G.

Banaji

Mahzarin R.

2005. Understanding and using the Implicit Association Test: II. Method variables and construct validity. Personality and Social Psychology Bulletin 31 (2): 166–80.

72.

Padavic

Irene

Ely

Robin J.

Reid

Erin M.

2020. Explaining the persistence of gender inequality: The work–family narrative as a social defense against the 24/7 work culture. Administrative Science Quarterly 65 (1): 61–111.

73.

Phelan

Julie E.

Moss-Racusin

Corinne A.

Rudman

Laurie A.

2008. Competent yet out in the cold: Shifting criteria for hiring reflect backlash toward agentic women. Psychology of Women Quarterly 32 (4): 406–13.

74.

Pritlove

Cheryl

Juando-Prats

Clara

Ala-Leppilampi

Kari

Parsons

Janet A.

2019. The good, the bad, and the ugly of implicit bias. The Lancet 393 (10171): 502–504.

75.

Rayner

Gerry M.

Papakonstantinou

Theo

. 2015. Employer perspectives of the current and future value of STEM graduate skills and attributes: An Australian study. Journal of Teaching and Learning for Graduate Employability 6 (1): 100–15.

76.

Reuben

Ernesto

Sapienza

Paola

Zingales

Luigi

. 2014. How stereotypes impair women’s careers in science. Proceedings of the National Academy of Sciences of the United States of America 111 (12): 4403–4408.

77.

Rice

Lindsay

Barth

Joan M.

2016. Hiring decisions: The effect of evaluator gender and gender stereotype characteristics on the evaluation of job applicants. Gender Issues 33 (1): 1–21.

78.

Rice

Lindsay

Barth

Joan M.

2017. A tale of two gender roles: The effects of implicit and explicit gender role traditionalism and occupational stereotype on hiring decisions. Gender Issues 34 (1): 86–102.

79.

Richman

Laura Smart

van Dellen

Michelle

Wood

Wendy

. 2011. How women cope: Being a numerical minority in a male-dominated profession. Journal of Social Issues 67 (3): 492–509.

80.

Rudman

Laurie A.

Fairchild

Kimberly

. 2004. Reactions to counterstereotypic behavior: The role of backlash in cultural stereotype maintenance. Journal of Personality and Social Psychology 87 (2): 157–76.

81.

Sassler

Sharon

Glass

Jennifer

Levitte

Yael

Michelmore

Katherine M.

2017. The missing women in STEM? Assessing gender differentials in the factors associated with transition to first jobs. Social Science Research 63:192–208.

82.

Seron

Carroll

Silbey

Susan

Cech

Erin

Rubineau

Brian

. 2018. “I am not a feminist, but. . .”: Hegemony of a meritocratic ideology and the limits of critique among women in engineering. Work and Occupations 45 (2): 131–67.

83.

Singh

Romila

Zhang

Yejun

Wan

Min Maggie

Fouad

Nadya A.

2018. Why do women engineers leave the engineering profession? The roles of work–family conflict, occupational commitment, and perceived organizational support. Human Resource Management 57 (4): 901–14.

84.

Smeding

Annique

. 2012. Women in science, technology, engineering, and mathematics (STEM): An investigation of their implicit gender stereotypes and stereotypes’ connectedness to math performance. Sex Roles 67 (11–12): 617–29.

85.

Stamarski

Cailin S.

Son Hing.

Leanne S.

2015. Gender inequalities in the workplace: The effects of organizational structures, processes, practices, and decision makers’ sexism. Frontiers in Psychology 6:1400.

86.

Tajfel

Henri

Turner

John C.

1979. An integrative theory of intergroup conflict. In The social psychology of intergroup relations, edited by Austin

William G.

Worchel

Stephen

. Pacific Grove, CA: Brooks/Cole.

87.

Tati

Firman

Riandi

2017. The effect of STEM learning through the project of designing boat model toward student STEM literacy. Journal of Physics: Conference Series 895:012157.

88.

Thébaud

Sarah

Charles

Maria

. 2018. Segregation, stereotypes, and STEM. Social Sciences 7 (7): 111.

89.

Thelwall

Mike

Abdullah

Abrizah

Fairclough

Ruth

. 2022. Researching women and men 1996–2020: Is androcentrism still dominant? Quantitative Science Studies 3 (1): 244–64.

90.

Turner

John C.

Hogg

Michael A.

Oakes

Penelope J.

Reicher

Stephen D.

Wetherell

Margaret S.

1987. Rediscovering the social group: A self-categorization theory. Oxford: Basil Blackwell.

91.

Uhlmann

Eric Luis

Cohen

Geoffrey L.

2005. Constructed criteria: Redefining merit to justify discrimination. Psychological Science 16 (6): 474–80.

92.

Van Dalen

Hendrik P.

Henkens

Kène

Schippers

Joop

. 2010. Productivity of older workers: Perceptions of employers and employees. Population and Development Review 36 (2): 309–30.

93.

Wang

Ming-Te

Degol

Jessica L.

2017. Gender gap in science, technology, engineering, and mathematics (STEM): Current knowledge, implications for practice, policy, and future directions. Educational Psychology Review 29 (1): 119–40.

94.

Weisgram

Erica

Diekman

Amanda

. 2015. Family friendly STEM: Perspectives on recruiting and retaining women in STEM fields. International Journal of Gender, Science and Technology 8 (1): 38–45.

95.

Williams

Joan C.

2004. Hitting the maternal wall: Before they reach a “glass ceiling” in their careers, women faculty may hit a “maternal wall.” Academe Bulletin of the American Association of University Professors 90 (6): 16–20.

96.

Williams

Wendy M.

Ceci

Stephen J.

2012. When scientists choose motherhood: A single factor goes a long way in explaining the dearth of women in math-intensive fields. How can we address it? American Scientist 100 (2): 138–45.

97.

Wynn

Alison T.

Correll

Shelley J.

2018. Combating gender bias in modern workplaces. In Handbook of the sociology of gender, edited by Risman

Barbara

Froyum

Carissa

Scarborough

William J.

Cham, Switzerland: Springer.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB