Before the experiment, participants completed online measures of chronic perceptions of sexism, anxiety, depression, and personal control as part of a larger set of questionnaires.
We modeled the measure of chronic perceptions of sexism after that used byKaiser et al. (2006)
. Women indicated their agreement—on a scale from 0 (strongly disagree
) to 6 (strongly agree
)—with five statements: “In general, others respect the gender group I am a member of” (reverse scored), “I experience discrimination because of my gender,” “My gender group is discriminated against,” “I have been unfairly treated because of my gender,” and “Members of my gender group face a good deal of gender discrimination.” The items were averaged to form a highly reliable scale α = .89 (M
= 3.23, SD
To control for individual differences that may covary with chronic perceptions of sexism, we also assessed participants’ anxiety, depression, and perceptions of personal control. We measured anxiety and depression using items from the Brief Symptoms Inventory (Derogatis & Spencer, 1982
). Participants reported how often they felt five anxiety symptoms (e.g., nervousness or shakiness inside, tense or keyed up; α = .82) and six depression symptoms (e.g., lonely, hopeless about the future; α = .86) on a scale from 0 (never
) to 6 (all of the time
). We measured personal control with a five-item version of Pearlin and Schooler’s (1978)
Mastery Scale. This measure contains items such as “What happens to me in the future mostly depends on me” (α = .63). Participants reported how true each statement was of them on a scale of (not at all true
) to 6 (very much true
We modeled our procedure on published studies examining dyadic interactions using the “workgroup paradigm” (e.g., Major et al., 2002
) and followed standard procedures for collecting salivary cortisol (Kirschbaum & Hellhammer, 1989
). Participants were run between 2:00 p.m. and 7:00 p.m. when cortisol levels are at their waking nadir. One day in advance of their scheduled lab visit, participants were sent email guidelines asking them to refrain from activities that could influence their cortisol levels. The list included, for example, brushing their teeth or drinking caffeinated beverages within 4 hours of their scheduled time and exercising at any point on the day of their scheduled time.
Women arrived individually but waited outside of the laboratory with a male confederate. The experimenter greeted them and said that even though they were waiting for a third “participant”—Andy (sexist condition) or Amy (merit condition)—they would get started with the study. The experimenter then escorted the participant into the experimental room while a second experimenter led the confederate into another room. Participants read that the study involved measuring the body’s stress response during interview situations. They then completed the hormone screening form followed by a 14-item bogus leadership questionnaire designed to bolster our cover story and provide a merit-based reason for feedback in the merit condition. After participants had been in the lab 20 minutes, they provided a baseline saliva sample.
Next, the participant and both of the confederates introduced themselves via video. Over an intercom, the experimenter told them to give their first name, year in school, major, and hobbies. Participants saw the confederate whom they had seen in the hallway introduce himself first, followed by the “late” bogus participant, Amy or Andy. Both introductions were pretaped and showed the confederates in an experimental room similar to the participants’. Participants introduced themselves last.
The experimenter then explained that based on a random drawing, the participant and Amy or Andy were assigned roles as applicants and the other, male, confederate was assigned the role of interviewer. His job was to interview the applicants and choose one to be his partner on a task with a chance to win $50. He was then given time to review both applicants’ performances on the leadership questionnaire, which were ostensibly scored using “a system developed by the Stanford Graduate School of Business.” The experimenter then asked him to state his initial impression of the two applicants over the intercom.
This audio feedback was pretaped and was negative in both conditions. Participants in the merit condition heard, “Uh . . . well, it says here that Participant B got a lower score on the leadership questionnaire . . . so, just based on that, I’d have to say that at this point I would pick Participant A over B. I think I will have a better chance of winning the prize if I work with her.” Participants in the sexist condition heard, “Uh . . . well, Participant B is probably too emotional and won’t be a strong partner . . . so, I’d have to say at this point that I would pick Participant A over B. I think I will have a better chance of winning the prize if I work with him.” Thus, the only difference between conditions was the reason given for potential rejection.
Participants were then given 2 minutes to prepare for the interview. They were told that during the interview they should try to convince the interviewer to select them by explaining why their talents, skills, and experiences make them the best teammate.
The experimenter then brought the confederate into the participant’s experimental room and seated him at a table facing her. To maintain the cover story, the experimenter explained that the participant would be interviewed prior to the second applicant. The experimenter left the room and then, over the intercom, instructed the participant to begin speaking. If she stopped speaking before the required 5 minutes had elapsed, the experimenter prompted her to continue.
Next, the participant completed a 5-minute backward digit span task while the interviewer kept track of her performance. Over the intercom, participants heard a prerecorded adult female voice recite a list of 19 sets of two-digit numbers (4 to 6 numbers per set). Their task was to repeat the numbers in reverse order immediately following each set. Following this, the experimenter escorted the confederate out of the room.
Subsequently, the participants reported how stressful they thought the interview was and then completed the manipulation check. We collected saliva samples at three additional times, 20, 30, and 40 minutes following the stressor, which was operationalized as when participants heard the rejecting feedback. These times were selected based on the consistent finding that peak concentrations of salivary cortisol occur about 20 minutes following a stressor (Dickerson & Kemeny, 2004
). Finally, participants were probed for suspicion concerning the feedback and confederates and were then fully debriefed and, when applicable, paid.
We collected saliva samples by having participants expectorate 1 ml of saliva into IBL (Hamburg, Germany) SaliCap sampling devices using a plastic straw. SaliCaps were stored in a −20°C freezer until shipped on dry ice to be assayed. Saliva samples were assayed at the California National Primate Research Center at the University of California, Davis. Prior to assay, samples were centrifuged at 3,000 rpm for 20 minutes to separate the aqueous component from mucins and other suspended particles. Salivary concentrations of cortisol were estimated in duplicate using commercial radioimmunoassay kits (Diagnostics Products Corporation, Los Angeles, CA). Intra- and interassay coefficients of variation were 2.44 and 3.97, respectively.
We created three measures of cortisol reactivity by subtracting baseline cortisol values from the cortisol values at Time 2 (20 minutes poststressor), Time 3 (30 minutes post-stressor), and Time 4 (40 minutes poststressor). Higher values indicate greater cortisol. Since cortisol levels naturally decline from awakening to afternoon (Schmidt-Reinwald et al., 1999
) and our procedure was less stressful than those that typically show increases in cortisol (e.g., Dickerson & Kemeny, 2004
), we expected a drop in cortisol levels over the course of the experimental session for all participants. Consequently, we focus on differences in levels
of cortisol, not on increases per se (e.g., Miller & Maner, 2009
; Page-Gould, Mendoza-Denton, & Tropp, 2008
). To maintain consistency with the literature, we use the term cortisol reactivity
to refer to our measure of change in cortisol from baseline (e.g., Dickerson & Kemeny, 2004
; Page-Gould et al., 2008