summarizing 2 short articles both 30 pages then comparing between both in an essay

https://doi.org/10.1177/0306312718756766

Social Studies of Science 2018, Vol. 48(1) 149 –164

© The Author(s) 2018 Reprints and permissions:

sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0306312718756766

journals.sagepub.com/home/sss

Puncturing the pipeline: Do technology companies alienate women in recruiting sessions?

Alison T Wynn The Clayman Institute for Gender Research and Department of Sociology, Stanford University, Stanford, CA, USA

Shelley J Correll Department of Sociology and Clayman Institute for Gender Research, Stanford University, Stanford, CA, USA

Abstract A ‘chilly’ environment limits women’s advancement through the educational pipeline leading to jobs in science and technology. However, we know relatively little about the environment women encounter after making it through the educational pipeline. Do technology companies create environments that may dampen women’s interest at the juncture when they are launching their careers? Using original observational data from 84 recruiting sessions hosted by technology companies at a prominent university on the US West Coast, we find that company representatives often engage in behaviors that are known to create a chilly environment for women. Through gender-imbalanced presenter roles, geek culture references, overt use of gender stereotypes, and other gendered speech and actions, representatives may puncture the pipeline, lessening the interest of women at the point of recruitment into technology careers.

Keywords gender, STEM, stereotypes, work and occupations

Tech goes to university: A typical recruiting session

You walk into the room and you’re immediately greeted by a bubbly female recruiter. ‘Welcome!’, she says enthusiastically. ‘Please sign in and take a raffle ticket. We’ll be auctioning off an iPad! And there’s plenty of food in the back!’

Corresponding author: Alison T Wynn, Department of Sociology, Stanford University, MC 2047, 450 Serra Mall, Stanford, CA 94305-2047, USA. Email: [email protected]

756766 SSS0010.1177/0306312718756766Social Studies of ScienceWynn and Correll research-article2018

Research Note

150 Social Studies of Science 48(1)

At the front of the room, a man is preparing his notes and setting up the projector screen. (Perhaps there are two men, or a whole cluster.) These are the engineers, the content presenters.

A few minutes pass, and the female recruiter asks the presenter, ‘Should we go ahead and get started?’ The man nods his approval.

He kicks off the presentation by introducing himself. (He rarely introduces the female recruiter.) Then he asks the audience questions about their background. ‘How many of you are CS majors? Double E? What else?’ The students murmur their answers.

After these questions, he briefly quizzes the audience about their company knowl- edge. ‘How many of you have heard of [insert company name here]? What are some of our products you’ve heard of?’ A student or two (usually men) shout out some answers.

The presenter begins the formal presentation by explaining what the company does at a high level. The slides almost certainly contain some geek culture references – Star Trek is a popular choice – and likely contain pictures of stereotypically masculine men to represent workers in the company. Then the presenter launches into a discussion so bru- tally technical that the students’ eyes glaze over; you can see many of them on their computers or whispering to one another, and sometimes they leave the session early. The diagrams on the slides have arrows pointing in confusing directions, the text is too small to read, and the speech is full of jargon.

If the company sends a female engineer, often she has no speaking role at all. If she does speak, her segment centers on ‘company culture’ or the company’s vision. Other women recruiters are in charge of handing out swag and other materials, or managing logistics. The ‘hard-core’ technical material belongs to the male presenters.

In the company culture segment, the presenter mentions the ‘fun’ work environment, with a plethora of ‘perks’, such as free meals and haircuts, which mean employees never need to leave the office. Entertainment might include foosball tables, beer fridges, and other elements of fraternity culture. Sleep is a low priority, mentioned as a point of pride.

Then the main male presenter opens the floor for a Q&A segment. Only male students ask questions. The conversation becomes a chest-beating competition – who can ask the most technical question that stumps the presenter? Meanwhile, the few female students remaining in the room stay silent.

The leaky pipeline leading to tech jobs

What happens when women with science, technology, engineering, and mathematics (STEM) degrees encounter firms interested in hiring them? In this article, we analyze observational data from 84 recruiting information sessions hosted by technology compa- nies at a US West Coast university known for its large concentration of engineering and computer science majors. We note the gender representation and roles of the company representatives, the images shown on their slides, videos and swag that they give to attendees, the comments by presenters and their interactions with the audience, and the reactions of student attendees. We are interested in the ‘chilly climate’ that women expe- rience in STEM fields, a climate that contributes to the continued dearth of women in these fields (e.g. Cech et al., 2013; Faulkner, 2009).

Wynn and Correll 151

To provide a feel for the flow of these sessions from the time students enter until the session disperses, we started with a hypothetical recruiting session narrative based on data from our observations. While there is variation from session to session, the narrative above is meant to provide a rough guide to elements that were common to many sessions.

While 40 percent of men with STEM college degrees work in STEM jobs, only 26 percent of women do (Beede et al., 2011). Even as women’s share of the college-edu- cated workforce has increased over the past decade, women’s underrepresentation in STEM fields has remained relatively constant or even decreased. This disparity has important implications for lifetime earnings; women in STEM jobs earn 20 percent more than do comparable women in non-STEM jobs (Beede et al., 2011).

Much past research has shown how educational experiences diminish young women’s interest in pursuing STEM fields. Other studies, which we review below, have relied on social psychological laboratory experiments to study recruitment-like scenarios, but these studies are often criticized for creating ‘artificial’ environments that do not match real-world settings. By observing a setting researchers typically do not access – one key to companies’ hiring – our research provides a window into social phenomena heretofore not observed and highlights a mechanism by which companies ostensibly trying to recruit women instead deflate their interest in technology careers.

According to social psychological research, certain environments can lower wom- en’s interest and sense of belonging in STEM fields, thereby contributing to the dearth of women in these fields. Factors that can make an environment chilly for women are: sexually explicit references, stereotypic images, references to cultural knowledge, hobbies and behaviors that are culturally associated with men and masculinity, being excluded from conversations or having one’s contributions ignored, and low numbers of women (e.g. Cheryan et al., 2009; Faulkner, 2009; Kanter, 1977; Murphy et al., 2007; Turco, 2010).

Viewing gender-stereotypic commercials (Davies et al., 2002) and geeky images such as Star Trek posters and video game boxes (Cheryan et al., 2009) lowered wom- en’s interest, performance, and sense of belonging in STEM fields. Men did not expe- rience a similar effect. Thus images in the physical environment can evoke or impede women’s sense that they belong, ultimately influencing their interest in pursuing STEM domains.

Researchers also find that heavy emphasis on prior technical experience or high skill level can make women feel unwelcome even when they are just as qualified as their male counterparts (Margolis and Fisher, 2002). Women tend to hold themselves to a higher standard than men in male-dominated fields such as math and engineering, and women require more evidence of their ability in order to feel competent (Correll, 2001, 2004) Environments that emphasize multiple successful paths into STEM fields are more effec- tive in attracting women (Margolis and Fisher, 2002).

Additionally, the role models presented to students can influence their interest in STEM careers. Murphy et al. (2007) find that women majoring in a STEM field expressed less of a sense of belonging, were less interested in attending a hypothetical engineering conference, and had greater increases in heart rate and skin conductance when they viewed a gender-unbalanced advertising video (three men to one woman) compared to a gender-balanced video (one man to one woman).

152 Social Studies of Science 48(1)

In studies such as these, researchers have identified the types of environments and role models that influence women’s interest in pursuing STEM careers. Even for highly qualified, motivated women majoring in STEM fields, priming stereotypes through the use of gendered or geeky images can have a significant negative effect on reported inter- est. Therefore, if companies reproduce these environments and behaviors in their recruit- ing sessions, they may discourage women from applying. Tech companies overwhelmingly lack women in high positions, and they tend to host a masculine-geeky ‘brogrammer’ culture that can alienate existing women employees and potential job candidates (Faulkner, 2007, 2009; Gross, 2012; Oldenziel, 1999; Varma, 2007; Wajcman, 1991, 2004). Additionally, tech companies often valorize long work hours and a youthful, fam- ily-free devotion to work (Cooper, 2000; Frank, 2013), which is potentially more alienat- ing for women than for men.

Data and methods

Each autumn, companies ranging from small start-ups to large multinational corpora- tions flock to our research site, to hold informational recruiting sessions with the goal of enticing undergraduate and graduate students to apply for technical positions. Many tech companies believe they are engaged in a ‘talent war’ for ‘top talent’ and therefore lav- ishly recruit at a few elite universities, including our research site (Michaels et al., 2001). Food and swag are routinely handed out in order to encourage students to attend the ses- sions and steer this ‘talent’ toward their company. These sessions precede companies’ on-campus interviews as an effort to generate student awareness and enthusiasm.

In 2012 and 2013, with the help of three undergraduate research assistants, we attended a total of 84 of these technical recruiting sessions from 66 different companies. Our goal was to observe the extent to which these sessions paint an image of tech work as ‘chilly’ to women. Based on past research identifying factors that dampen women’s interest or sense of belonging, we define an environment as ‘chilly’ if, through numerical representation and gendered roles, it conveys a sense that men are more prevalent or more important. An environment can also be chilly if it is permeated with references to cultural knowledge, hobbies, or behaviors that are more commonly associated with men than women, or if women are referred to in sexualized or other gendered ways. If tech recruiting sessions are chilly environments, they potentially dampen the movement of women into tech work.

Three female undergraduate research assistants assisted the first author in making the observations. This research team was young enough to blend in with the students in the sessions. We located tech sessions by viewing the online event calendars for the univer- sity’s career center and computer science department. We completed two distinct rounds of observation. In 2012, we observed twenty pilot sessions using an open style of obser- vation, taking detailed notes and allowing themes to emerge inductively from the data. In 2013, we completed a more systematic round of observation based on a thorough reading of existing literature and the emergent themes from the 2012 data. Researchers used a detailed observation guide in each session to note key patterns, and we observed 64 ses- sions from 59 unique companies using this focused style of data collection. (A copy of the observation guide is available upon request.) Since our goal was to observe elements

Wynn and Correll 153

of the environment that might be chilly, the observation guide included categories such as presenter information, student demographics, references to potentially gendered expe- riences, valued traits, references to stereotypically ‘geeky’ items, unnecessary/excessive technical complexity, work/life balance references, role models discussed, slide and video content, ‘swag’ details, gendered rhetoric/speech, gender-neutral examples, gen- eral tone, and Q&A details. Each category included a detailed description, which pro- vided specific examples from existing research. Research assistants were trained prior to observation, and we initially attended one session together to compare and calibrate our note taking. After this calibration, each session was observed by one member of the research team. We attended these sessions as a ‘fly on the wall’, observing and taking notes without influencing or interacting with the students or company representatives.1

Each informational recruiting session lasted one hour. Attendance ranged from eleven to about 100 students, with an average of 33 students.2 On average, twenty-four percent of the students were women; this ranged from zero women in one session to fourteen women out of nineteen students (74 percent) in another session.3 Students were typically computer science or electrical engineering majors; however, other majors were also rep- resented (e.g. MBA students, other math or engineering majors, etc.). Interestingly, com- puter science is one of the most popular majors on this campus for both women and men. Company representatives discussed a range of topics in the sessions, from company products and services to company culture and available positions. We use pseudonyms throughout, and identifying information about companies has been stripped from our reported results. The examples we present are representative of key patterns in the data.

Major themes

The typical narrative with which we began this article depicts some of the major themes and patterns that emerged during our observations. We observed four main themes: the presentation and discussion excluding women, the pervasive use and enforcement of gender stereotypes in content, extreme technicality, and references to masculine geek culture.

Many of these patterns occur when presenters attempt to be personable and relata- ble to their student audience. As they make a genuine attempt to connect with the audi- ence on a personal level, presenters often convey a gendered sense of self. Since the presenters are primarily male, acting personable often means enacting masculinity, as we show below. Because men presenters represent not only themselves, but also their companies, their masculine enactments become emblematic of a ‘fratty’ or a ‘geeky’ company culture, suggesting who is likely to fit in, which may deter qualified women from applying.

Many of the patterns we describe below vary with company size and age. Larger, more established companies seem more aware of the image they project, and they tend to have more formal, less gendered presentations. Presenters focus more on communicat- ing facts about the company, rather than joking around with the audience. Many different factors may combine to create this overall effect; larger companies may have more for- malized human resources departments that more closely vet the presentation scripts. These companies could also have more women on staff, or they could have less

154 Social Studies of Science 48(1)

connection to the notorious ‘start-up culture’ of Silicon Valley (Cooper, 2000; Gross, 2012). Whatever the reason, size does appear to matter when it comes to the chilliness of recruiting sessions.

Theme #1: The presentation and discussion exclude women

Presenter roles: Expert men, marginal women. As indicated in the typical narrative, men typically presented the bulk of the technical content. Women company representatives tended to be marginalized or even silent. Eighty-four percent of the sessions followed this pattern of primary male presenters. Twenty-two percent of the sessions had no women presenters at all, 58 percent had women in marginalized roles, and in four percent of the sessions, though they presented technical content women clearly played a smaller role than their male counterparts. Only 22 percent of the sessions had female engineers presenting core technical content.

Across many sessions, male presenters shared the spotlight with one another, while female company representatives set up the food and decorations, handed out t-shirts, and collected resumes, often without ever speaking or being introduced to the audience by their male colleagues. For example, in a start-up company’s session with three male soft- ware engineers and two female recruiters, a male presenter urged the students to pass their raffle tickets down to the ‘lovely ladies’. In another session, a woman recruiter was intro- duced in a particularly gendered way: ‘This is Kathy. She’s really nice. She cries easily.’

Sometimes women resisted their marginalized role in small but perceptible ways. One large, established company’s session started out following the typical pattern: A female recruiter, Jane, set up the food and handed out the t-shirts before the session began. During the presentation, while her male colleague Kevin talked at length, Jane stood in the back of the room, completely hidden behind the wall of students in the audience, literally invisible. At the end of the session, Jane emerged from behind the students. ‘I didn’t get a chance to introduce myself at the beginning’, she said. After introducing herself as the university recruiter, she explained to the students the logistics of how to apply for a job.

In an internet start-up session, two men (Ramit and Lucas) shared the presenting, while the woman, Mei, remained silent. At the end, Ramit said, ‘Let me bring Mei up before we take questions.’ He introduced her – she did not introduce herself – and told the audience she had been working at the company since August and got her PhD at the same university attended by the audience. Though Lucas and Mei had the same technical position at the company, Lucas had a much larger role in the presentation. Then they began the Q&A segment, which became a male-dominated conversation that almost entirely boxed Mei out. Ramit answered all nine questions (asked by four male students), and Lucas weighed in for three of the questions. Mei tried to jump in for two questions, but Ramit cut her off at one point and directed attention to Lucas, saying, ‘He [Lucas] manages a team.’ At the very end, Mei said, ‘I want to add something’ and talked about the real-world impacts of their work. Throughout the Q&A, Mei had to fight for her right to speak, the two men routinely interrupting her.

Thus, even when companies brought women representatives, they often had margin- alized roles. In many sessions, the women didn’t speak at all, and if they did speak they

Wynn and Correll 155

tended to cover ‘softer’ topics, such as social aspects, work-life balance, company cul- ture, and logistics of applying. In addition, the men frequently interrupted the women presenters. Large, well-known companies as well as small start-ups followed these pat- terns. Through selection of role models and delegation of tasks, companies sent a subtle message to potential candidates about organizational culture and demographics.

Male-dominated Q&A segments. At the end of each session, presenters invited students to ask questions. Sometimes these Q&A segments lasted for half the session, and other times the presenters answered only one or two questions at the very end. The sessions also ranged in formality: Sometimes presenters neatly called on one student at a time, while in other sessions, the Q&A became more like a back-and-forth conversation. In all cases, the male students asked the vast majority of the questions. They also tended to use Q&A segments as an excuse to compete, show off their technical expertise, and com- mandeer the discussion. A few outspoken men tended to dominate the conversation, sometimes asking three or four questions each. The company representatives did not effectively facilitate or manage this dynamic; instead they were often pulled into the discussion, joining into the competition and volleying back and forth with the vociferous male students.

For example, in one well-known company’s session, the men asked nineteen ques- tions (with some men asking more than one question), while the women asked none. Four of the fifteen women in attendance left during the Q&A, whereas only one of the 51 men left. Of the five presenters, the two male engineers answered the vast majority of the time. The female recruiter jumped in toward the end to answer questions about the recruiting process, but the other two women engineers rarely spoke. In fact, they looked rather bored; they stood off to the side, and one woman gazed out the window, arms crossed. The (male) students asked highly technical questions, and the (male) presenters responded in kind. In another session, five men asked fifteen questions, and one man asked nine of the questions. In this way, many companies’ Q&A sessions devolved into a highly technical competition between a few men, leaving the women (and most men) silent and stranded. Such Q&A sessions paint an image of a competitive company culture that many students – particularly women – may not seem to fit.

Theme #2: The pervasive use and enforcement of gender stereotypes in content

Overt use of gender stereotypes. Company representatives also occasionally showed slides containing overt gender stereotypes, such as objectified or sexualized images of women. For example, one sub-group of a large online retailer posted a slide with Popeye and Olive, characters from an old cartoon. Popeye, a man flexing his burly biceps, had the label ‘data plane’, and Olive, his demure female counterpart – her heart shown beating out of her chest, as if she were swooning over her man – had the label ‘control plane’. The male presenter explained: ‘You have a big beefy data plane, and a slender control plane.’

A start-up gaming company showed an animated woman wearing a red, skin-tight dress and holding a burning poker card in her hand to represent a poker game product. Another company exclusively posted images of men as the active subjects – astronauts,

156 Social Studies of Science 48(1)

computer technicians, and soldiers shooting each other with powerful guns in a violent video game. The presenter even showed a picture of a man raising his hand when encour- aging the students to ask questions. Only one slide depicted a woman, when demonstrat- ing how the company’s product can help users construct graphic images. The virtual woman struck a seductive pose, looking over her shoulder suggestively, and thus served as a sexualized object to be created rather than an active subject. Such imagery may make gender salient for students in the room and subtly tell women that they do not belong in technical environments.

In addition to displaying objectified images of women, companies also occasionally made gendered pop culture references, most notably in an attempt to relate to students. For example, Felix, CEO and co-founder of an ad-targeting company, used the show Mad Men to frame his entire presentation. An early slide showed an image of Don Draper, the male protagonist, staring confidently into the camera. ‘This is Don Draper’, Felix explained. ‘He has the most important and best paid job [in the show’s advertising agency].’ The following slide showed ‘his counterpart – Harry Crane. His job is so easy’, according to Felix. ‘He spends most of the day watching TV.’ Felix used this comparison to explain how advertising has changed since the 1960s, now that television channels have exploded and new media options abound. ‘Today’s Harry Cranes are using com- puter science’, Felix explained later in the talk. Not only is Mad Men famous for exem- plifying the rampant sexism of 1960s advertising agencies, but by portraying two men and no women in his slides, the presenter conveys that while today’s Harry Crane might use computer science, he is still a man.

Similarly, in a major computer company’s presentation, a male presenter said, ‘As you can see from the Tommy Bahama’s photo, I am the most interesting man in the world’, a reference to a masculinized Dos Equis beer commercial. An internet start-up showed an image of the Gangnam Style music video, featuring male artist Psy sur- rounded by scantily clad women. Companies seem to provide these gendered pop culture references to seem funny or cool, but in doing so they also convey a gendered sense of who fits in their company culture.

Fratty environment. Company representatives also offered commentary that indicated a ‘fratty’, young, masculinized work environment. For example, one of our female under- graduate research assistants wrote the following about a start-up founded by a group of graduates from the university where the data were collected:

They tried to make the work environment seem fun, though it came across as stereotypically masculine. Some things they advertised on the slides were snacks, nerf gun army, scooters, catered lunch, ping-pong, foosball, pool, beer pong table, and beer fridges. The last few are usually associated with [our university’s] frat culture, so I imagine they may come across that way to … students.

Since these company representatives were alumni from the same university, they were presumably familiar with the university’s fraternity culture and its cultural icons.

Later in the session, one male presenter, Dan, mentioned the ‘better gender ratio’ at the company’s Los Angeles office compared to its Silicon Valley office. However, rather

Wynn and Correll 157

than using this point as an opportunity to highlight women’s representation in the com- pany, Dan, who worked at the LA office, explained, ‘I had no girlfriends at [University name], but now I’m married.’ This portrays the ‘better gender ratio’ as an opportunity for men to court their women coworkers romantically.

Similarly, a gaming start-up company showed a short video that included numerous gendered comments. One man in the video boasted that he could ‘relieve some stress’ while on the job because of the onsite gym; as he spoke, he punched a punching bag. In contrast, a woman proclaimed: ‘Everyone I work with is a rockstar.’

Company representatives occasionally created this ‘fratty’ vibe by cursing or using other informal, joking language. In the gaming company’s session, the presenter uttered phrases like ‘they shit the bed’ and ‘it’s a real pain in the ass’. He even became aggressive at times, telling a male student he was ‘completely wrong’. At another session, the pre- senter said, ‘We’re from [University name] – we’re the shit.’ While not explicitly gen- dered, such language may create an environment that feels reminiscent of a fraternity-like culture.

Occasionally, this informal rhetoric devolved into outright sexualized comments. For example, one small tech start-up representative mentioned pornography twice. First, he talked about a vision for the future of handheld computers, ‘where they would gamble and look at porn’. Though he could have mentioned any number of uses for handheld computers, he referred to two gendered activities. Later in the presentation, he mentioned pornography again when answering a female student’s question; the student reacted with visible shock and discomfort. Similarly, one presenter from a well-known company men- tioned prostitution. When describing a project that involved developing a banking sys- tem on ships, he listed reasons why sailors needed access to cash when they reached port. ‘I won’t go into the other things sailors do at port’, he jibed, ‘but they needed cash.’ He spoke in a suggestive tone. Using such informal, ‘dirty’ language may convey a mascu- line, one-of-the-guys atmosphere that is potentially alienating to women.

Gendered swag. Companies also sometimes gave away t-shirts or other ‘swag’ with gen- dered messages. One company handed out t-shirts that read, ‘I like big data and I cannot lie’, playing off the well-known song lyric ‘I like big butts and I cannot lie.’ Another t-shirt read, ‘I’ll show you my data if you show me yours.’ One company distributed black notebooks that said, ‘Finding your faults, just like mom.’ Two other companies gave away beer products such as bottle openers and beer steins. Smaller start-ups in par- ticular were most likely to hand out gendered swag. These tangible gifts provide a sense of the company’s culture, values, and prevailing gender norms.

Theme #3: Extreme technicality

While some technical complexity is necessary to communicate the requirements of the job, many presenters seemed to use technical complexity as a method of vetting or intim- idating the audience. Because women tend to have less confidence in their technical skill – even when just as qualified as their male counterparts – heavy emphasis on over- whelming technical complexity may discourage women from applying (Correll, 2001, 2004; Margolis and Fisher, 2002).4 In fact, overwhelming or incomprehensible

158 Social Studies of Science 48(1)

complexity may discourage men from applying too, but the effects may be more pro- nounced for women.

For example, one large, well-established company stood out in its clear use of incom- prehensible, excessive complexity. The slides had diagrams too complicated for the human eye to follow, with text too small for the audience to read. The diagrams were multi-layered and complicated, and the presenter left each slide up for only a few sec- onds. Even for seasoned tech gurus, the presentation would have been impossible to follow. The presenters used incessant jargon and acronyms, both in their speech and in slide text, without explaining what their terms meant. Over half of the women students at this particular session left the room left early (eight out of fourteen), and almost a third of men left early as well (twelve out of thirty-eight). While these displays of technicality appealed to some men, a large number of men and women seemed turned off by these behaviors, an outcome ostensibly at odds with the goal of recruiting.

Many sessions were not so extreme; they merely used frequent undefined jargon or had a few complicated graphics. For example, in one large internet start-up session, male presenter Kyle displayed a slide with extremely small text. A male student sitting beside the researcher murmured ‘Jesus’ under his breath when the slide went up. Kyle admitted that the text was small, calling it ‘a big block of JSON’, or JavaScript Object Notation, a type of data format. Its small font and massiveness as a collective conveyed the sense of a huge, daunting hunk of code, emblematic of what tech workers at this company might be expected to create. In these subtle ways, presenters may intimidate qualified students, particularly women. As past research has shown, narrow definitions of success in tech that emphasize only technical skills and not the real-world implications of those techni- cal skills can lower women’s confidence and interest in pursuing tech careers (Margolis and Fisher, 2002).

Theme #4: References to masculine geek culture and gendered lives

Geek culture references. According to existing social psychological research, references to stereotypically geeky items – such as Star Trek, video games, comics, and t-shirts with messages like ‘I code, therefore I am’ – can alienate women in computer science fields (Cheryan et al., 2009, 2011). Even seeing such items in an empty room can make women feel that they do not belong in that environment (Cheryan et al., 2009). Importantly, these images operate as cultural signals of who belongs. Even though plenty of women enjoy video games and Star Trek, these images are culturally associated with masculinity and convey a masculine work environment.

In the recruiting sessions we attended, presenters often referred to such geeky items or included images on their slides. Across the sessions, the following geeky references emerged: He-Man, Star Trek, Star Wars, Terminator, The Matrix, Game of Thrones, Lord of the Rings, Pokemon, Hitchhiker’s Guide to the Galaxy, Back to the Future, The Big Bang Theory TV-show, Starcraft, Magic the Gathering, World of Warcraft, video games in general, Legos, and ‘Nerdy but Lovable’, ‘Data Rockstar’, and ‘I dream in code’ t-shirts. Almost 40 percent of sessions included a geek reference of some kind. Newer, smaller start-up companies were more likely to include these references than well-estab- lished companies.

Wynn and Correll 159

Typically, presenters used these references to seem funny or relatable to students. For example, one presenter showed a slide with a male Star Trek character saying, ‘WTF are you saying that I’m half Klingon?’ While not all male students laughed at this joke, the only people who did laugh were men. Presenters also used these references to demon- strate the company culture. A mobile start-up company displayed a slide labeled ‘Company of Gamers’, and the male presenter listed the video games they play together, even though video games have no relevance to the company’s product. ‘We don’t take ourselves too seriously, and we have a lot of fun with code’, he explained. These state- ments convey a gendered depiction of the company’s culture as a geeky-masculine envi- ronment in which women may expect they won’t fit.

References to potentially gendered experiences. Research indicates that women in technical professions often come to computing later in life than men and have less hands-on expe- rience with computing before entering college (Margolis and Fisher, 2002). Thus women may lack the early childhood experiences with computers, video games, and other tech- nology common among men in their field. Importantly, this differential in early experi- ence does not predict future success in the field (Margolis and Fisher, 2002). However, when presenters refer to such early experiences, they risk alienating the women in the audience by making them feel as though they don’t belong.

For example, one male representative of a large, established company said that employees gained most of their skills from playing video games as children. At another large company’s session, a male presenter said, ‘I get calls from my parents to fix their internet. I can tell from the nods that others have had this happen. How many of you have had this?’ Even though the women in the audience are engineers, their parents may have been less likely to approach them for computer help, due to stereotypes about women’s lack of competence in STEM fields (Margolis and Fisher, 2002; Steele, 1997). Some of these questions could be innocuous; however, when presenters choose items from child- hood experience, these remarks may make women feel like they don’t belong, since women may be less likely to have these early experiences.

References to lack of work-life balance. Companies frequently boasted of perks such as: free meals and drinks, catering and cafeterias, kitchens, Starbucks, gyms and fitness centers, yoga, climbing walls, Jr. Olympic sized pools, basketball courts, soccer and baseball fields, beach volleyball courts, ping-pong and billiards rooms, foosball tables, food trucks, on-site car washes and oil changes, dry cleaning, chiropractors and acu- puncturists, massages, haircuts, video games, social events and happy hours, speaker series, carnivals, bowling outings, scavenger hunts, poker tournaments, and ‘all-night hackathons’. While both men and women are likely to value these perks, company representatives occasionally stated explicitly that such perks mean never having to leave the office. According to one intern, ‘When I first moved here, I didn’t know where the USPS was, because everything was on campus.’ On-site chiropractors are ‘really great’ because ‘you can go for two minutes and come back [to work]’. Both work and play occur on the premises, and candidates with outside responsibilities – or candidates who expect to have outside responsibilities in the future – may find this implication problematic.

160 Social Studies of Science 48(1)

Other presenters stressed their devotion to long work hours. For example, Fredrick, a representative of a large company, repeatedly mentioned the long hours he devoted to work. Before the session began, he told a colleague, ‘I’ve been up since 6 am.’ When he began the session, he said, ‘I apologize. As I said, I haven’t eaten breakfast yet. I was falling over.’ At a start-up company’s session, one male presenter said of the other, ‘He forgot to go to bed, because he was having too much fun.’ He encouraged the students to visit the office ‘any time of day: 12 am, 2 am’, because employees will be there all times of day or night.

These examples demonstrate the ‘up-all-night’ start-up culture endemic to the Silicon Valley tech space (Cooper, 2000; Frank, 2013). One web start-up co-founder, Jared, showed pictures of his company in its early stages. ‘There was very little sleep’, with everyone working nonstop, according to Jared. He described this fondly as a ‘fun envi- ronment’. Start-up founders often express a glamorized view of late night sessions. But for students who prefer having a life outside of work or who anticipate having family responsibilities, this environment might seem more chilling than charming. And since women students worry more about family responsibilities than men do (Bass, 2015; Gerson, 2010), the valorization of being constantly present at work might especially deter women.

Positive examples

Past research provides some suggestions about what behaviors are likely to enhance women’s interest and sense of belonging. For example, emphasizing real-world, interdis- ciplinary impacts of technical coding has been shown to attract more women (Margolis and Fisher, 2002). Featuring more women engineers as role models (both as presenters in the room and as figures in videos and pictures) can also raise women’s interest (Murphy et al., 2007). Presenting technical work in an approachable, understandable way (for example, by defining jargon-filled terminology) can demonstrate that there are many acceptable pathways into computer science careers (Margolis and Fisher, 2002).

We observed these positive behaviors in about half of the sessions. For example, some companies made a concerted effort to detail and emphasize the real-world impacts of their code on global health, education, disaster response, life quality, and other important societal outcomes. In addition, women presented primary technical content in fourteen company presentations; the female Chief Operating Officer even presented the main content in one company’s session. A few companies explained their terms in an under- standable manner, and some companies had neutral slides focused on simple facts (such as market share, number of employees, etc.). One start-up company emphasized the importance of work-life balance; Jim, the company’s CEO, told the audience, ‘Let’s make a company; let’s not make a cult.’ Phrases on the slides read, ‘It’s a marathon, not a sprint’, ‘pace yourself, enjoy the journey’, and ‘family first’. Examples such as these provide an encouraging picture of what might be possible if companies make a concerted effort to welcome women. Women asked questions in 65 percent of the sessions with positive behaviors; in sessions without these positive behaviors, women asked questions only 36 percent of the time. Thus company representatives can have a profound impact on the engagement of the women in attendance at recruiting sessions.

Wynn and Correll 161

Conclusion

University recruiting sessions serve as an important stepping-stone to tech careers, and they provide a window into the processes by which companies and potential employees assess mutual fit. Past research has shown that assessments of fit often reproduce gen- dered patterns, with men appearing to fit more in male-dominated work environments (Kanter, 1977; Rivera, 2012; Sorensen et al., 2011; Turco, 2010). By observing tech recruiting sessions, we aim to understand whether and how companies convey what sort of person is likely to fit in their workplace. As we have shown, in a process that lacks clear guidelines or regulations, recruiters strive to be personable, relatable, and interest- ing to their student audiences. However, through their presentations, interactional styles, and the images they project, they convey a gendered sense of who will fit best in their company culture. By emphasizing geeky masculinity, they risk appealing only to a nar- row range of men and virtually no women. Smaller companies, in particular, compound this problem through the use of gendered swag, frequent references to geeky movies and TV-shows, and masculine cultural icons. The result is an environment that often feels like a fraternity house.

Paradoxically, most companies want to ‘win the talent war’ and claim that recruiting women is essential to that goal (Michaels et al., 2001). However, the cultural environ- ment they create interferes with that fundamental goal. While many company leaders point to the pipeline and related supply-side factors as their biggest constraint, our research suggests that companies may be partially responsible for the leaky pipeline car- rying women into tech careers.

More generally, supply-side processes, such as deciding to major in a particular field or apply for a particular job, are deeply intertwined with demand-side processes involv- ing the actions and behaviors of employers and other company representatives. Researchers often treat supply-side processes (e.g. women’s preferences, aspirations, and choices) and demand-side processes (e.g. decisions and judgments of gatekeepers, such as employers and teachers) as distinct conceptual categories. However, women’s decisions about what jobs to apply for, often considered an exogenous input to models testing for discrimination, may be, at least in part, the result of company behavior. Recruiting sessions provide an interesting site of observation in this regard since, during these sessions, companies explicitly present an image of their company culture and val- ues to potential new entrants. As such, recruiting sessions are a distinct point where companies and potential candidates come together to assess their mutual fit. The theo- retical implication of our findings is that researchers who study how individuals choose careers and make career-relevant decisions should be especially attentive to the environ- ment in which these decisions are made.

Our data provide a window into the actual recruiting process of companies. While our study contains detailed data on how company representatives convey a sense of company culture and some data on how many women and men participants ask questions or leave the session early, we lack direct evidence of women’s reactions to company presentations. However, prior literature has convincingly established that gender stereotypes, geek cul- ture references, excessive technical complexity, and lack of female representation system- atically decrease women’s interest in pursuing tech careers. Though we cannot know

162 Social Studies of Science 48(1)

definitively whether the women in our particular study had the same reactions, the recruit- ing environments we viewed were strikingly (almost shockingly) similar to the artificial environments created in social psychological experiments mentioned earlier.

On a practical level, our results suggest ways to create environments that are attrac- tive to a diverse range of women and men. Particularly in male-dominated fields, com- panies have an opportunity to increase the representation of women by being more mindful of the images they project in recruitment sessions, interviews, and the workplace more generally. By avoiding gender stereotypes, geek or fraternity culture icons, and other gendered references, company representatives may be able to increase the pipeline of women interested in tech jobs.

Acknowledgements

We would like to thank Lorna Bernhoft, Maggie Contreras, and Rachel Mewes for their role as Research Assistants. We would also like to thank the members of the social psychology workshop and the Michelle R. Clayman Institute for Gender Research at Stanford University for their helpful comments and suggestions.

Funding

This material is based upon work supported by the National Science Foundation Graduate Research Fellowship. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Notes

1. We did not attempt to collect data on students after the sessions. Doing so would have required us to inform students (and likely companies) about the purpose of our study. Since students in one session often appeared in subsequent sessions, revealing the purpose of our study could have altered their behavior in unrealistic ways, thereby interfering with our ability to observe these sessions. Even more importantly, it could have distracted students from their focus on the sessions themselves.

2. All mean and percentage calculations use the 2013 data exclusively, since we were not sys- tematically observing all sessions the same way in 2012. However, some examples may include data from 2012.

3. Race, ethnicity, and country of origin likely affect recruiting contexts and have important implications for minority representation in STEM fields. The vast majority of presenters and students in attendance appeared to be White or Asian, and so our sample of alternative races would not have been large enough to analyze race systematically. Further, since we did not interact with students, researchers would have had to assign race based on their own judg- ments, which has been shown to result in biased analyses and outcomes (Saperstein, 2006).

4. As the researchers possessed less background in technology than many of the students par- ticipating in recruiting sessions, estimating what was ‘excessively’ complex was not always a straightforward exercise. However, when companies reached particularly extreme levels of complexity, this was usually evident through the reactions of students (e.g. muttering under their breaths or leaving sessions early.) Obviously, technology companies should be expected to present technical topics in their recruiting sessions, but we attempted to note when com- panies did so in a ‘chest-beating’, intimidating, or overly masculine style, or in ways that could not be construed as giving students insights into the job requirements (e.g. when slides contained technical material using fonts that were not readable or when slides were presented too quickly to be read).

Wynn and Correll 163

ORCID iD

Alison T Wynn https://orcid.org/0000-0001-7518-9985

References

Bass BC (2015) Preparing for parenthood? Gender, aspirations, and the reproduction of labor market inequality. Gender & Society 49(3): 362–385.

Beede D, Julian T, Langdon D, et al. (2011) Women in STEM: A Gender Gap to Innovation (Report). Washington, DC: U.S. Department of Commerce, Economics and Statistics Administration.

Cech EA, Blair-Loy M and Rogers LE (2013) Recognizing chilliness: How cultural schemas of inequality shape views of culture and climate in work environments. In: Annual meeting of the American Sociological Association, New York, 10–13 August.

Cheryan S, Plaut VC, Davies PG, et al. (2009) Ambient belonging: How stereotypical cues impact gender participation in computer science. Journal of Personality and Social Psychology 97(6): 1045–1060.

Cheryan S, Siy JO, Vichayapai M, et al. (2011) Do female and male role models who embody STEM stereotypes hinder women’s anticipated success in STEM? Social Psychological and Personality Science 2(6): 656–664.

Cooper M (2000) Being the ‘go-to guy’: Fatherhood, masculinity, and the organization of work in Silicon Valley. Qualitative Sociology 23(4): 379–405.

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

Correll SJ (2004) Constraints into preferences: Gender, status, and emerging career aspirations. American Sociological Review 69: 93–113.

Davies PG, Spencer SJ, Quinn DM, et al. (2002) Consuming images: How television commercials that elicit stereotype threat can restrain women academically and professionally. Personality and Social Psychology Bulletin 28(12): 1615–1628.

Faulkner W (2007) ‘Nuts and bolts and people’: Gender-troubled engineering identities. Social Studies of Science 37(3): 331–356.

Faulkner W (2009) Doing gender in engineering workplace cultures, I: Observations from the field. Engineering Studies 1(1): 3–18.

Frank BH (2013) Evernote, salesforce CEOs: If you want work-life balance, work somewhere else. Geekwire, 27 September. Available at: http://www.geekwire.com/2013/marc-benioffs- advice-startup-founders-underestimate/ (accessed 16 December 2013).

Gerson K (2010) The Unfinished Revolution: How a New Generation Is Reshaping Family, Work, and Gender in America. New York: Oxford University Press.

Gross D (2012) In tech, some bemoan the rise of ‘programmer’ culture. CNN, 7 May. Available at: http://www.cnn.com/2012/05/07/tech/web/brogrammers/index.html (accessed 16 December 2013).

Kanter RM (1977) Men and Women of the Corporation. New York: Basic Books. Margolis J and Fisher A (2002) Unlocking the Clubhouse: Women in Computing. Cambridge, MA:

Massachusetts Institute of Technology. Michaels E, Handfield-Jones H and Axelrod B (2001) The War for Talent. Boston, MA: Harvard

Business School Press. Murphy MC, Steele CM and Gross JJ (2007) Signaling threat: How situational cues affect women

in math, science, and engineering settings. Psychological Science 18(10): 879–885. Oldenziel R (1999) Making Technology Masculine: Men, Women and Modern Machines in

America, 1870-1945. Amsterdam: Amsterdam University Press.

164 Social Studies of Science 48(1)

Rivera LA (2012) Hiring as cultural matching: The case of elite professional service firms. American Sociological Review 77(6): 999–1022.

Saperstein A (2006) Double-checking the race box: Examining inconsistency between survey measures of observed and self-reported race. Social Forces 85(1): 57–74.

Sorensen KH, Faulkner W and Rommes E (2011) Technologies of Inclusion: Gender in the Information Society. Trondheim: Tapir Academic Press.

Steele CM (1997) A threat in the air: How stereotypes shape intellectual identity and performance. American Psychologist 52(6): 613–629.

Turco CJ (2010) Cultural foundations of tokenism: Evidence from the leveraged buyout industry. American Sociological Review 75(6): 894–913.

Varma R (2007) Women in computing: The role of geek culture. Science as Culture 16(4): 359– 376.

Wajcman J (1991) Feminism Confronts Technology. University Park, PA: The Pennsylvania State University Press.

Wajcman J (2004) Technofeminism. Malden, MA: Polity Press.

Author biographies

Alison T Wynn is the Diversity & Inclusion Postdoctoral Fellow with the Clayman Institute for Gender Research at Stanford University. She received a PhD in sociology from Stanford University. Her research examines organizational polices and practices that may inadvertently create or rein- force inequality. In particular, she studies recruiting practices, perceptions of cultural fit, flexibility programs, and gender equality initiatives in industries such as technology, management consult- ing, and academic medicine.

Shelley J Correll is Professor of Sociology and, by courtesy, Organizational Behavior at Stanford University where she also serves as the Barbara D. Finberg Director of the Clayman Institute for Gender Research. Her current research examines how legal mandates and organizational policies and practices can reduce the effect of status beliefs on workplace gender inequalities and how gender status beliefs produce inequalities in product markets. She is currently conducting research in three technology companies to evaluate a ‘small wins’ model for reducing gender biases by involving organizational actors in the change process.