Gender, status, and psychiatric labels

Social Science Research 54 (2015) 68–79

Contents lists available at ScienceDirect

Social Science Research journal homepage: www.elsevier.com/locate/ssresearch

Gender, status, and psychiatric labels Amy Kroska a,⇑, Sarah K. Harkness b, Ryan P. Brown c, Lauren S. Thomas d a

Department of Sociology, University of Oklahoma, Kaufman Hall 331, Norman, OK 73019, United States Department of Sociology, University of Iowa, United States Department of Psychology, University of Oklahoma, United States d Cedar Park, TX, United States b c

a r t i c l e

i n f o

Article history: Received 23 June 2014 Revised 1 June 2015 Accepted 25 June 2015 Available online 29 June 2015 Keywords: Education Gender Modified labeling theory Status Stigma

a b s t r a c t We examine a key modified labeling theory proposition—that a psychiatric label increases vulnerability to competence-based criticism and rejection—within task- and collectively oriented dyads comprised of same-sex individuals with equivalent education. Drawing on empirical work that approximates these conditions, we expect the proposition to hold only among men. We also expect education, operationalized with college class standing, to moderate the effects of gender by reducing men’s and increasing women’s criticism and rejection. But, we also expect the effect of education to weaken when men work with a psychiatric patient. As predicted, men reject suggestions from teammates with a psychiatric history more frequently than they reject suggestions from other teammates, while women’s resistance to influence is unaffected by their teammate’s psychiatric status. Men also rate psychiatric patient teammates as less powerful but no lower in status than other teammates, while women’s teammate assessments are unaffected by their teammate’s psychiatric status. Also as predicted, education reduces men’s resistance to influence when their teammate has no psychiatric history. Education also increases men’s ratings of their teammate’s power, as predicted, but has no effect on women’s resistance to influence or teammate ratings. We discuss the implications of these findings for the modified labeling theory of mental illness and status characteristics theory. Ó 2015 Elsevier Inc. All rights reserved.

1. Introduction Psychiatric treatment programs can dramatically reduce patients’ symptoms (Link et al., 1997; Rosenfield, 1997). Nonetheless, studies over the past three decades show that this official recognition of mental illness is linked to declines in material, social, and psychological well-being (Kroska and Harkness, 2006, 2008; Link, 1982, 1987; Markowitz, 1998; Markowitz et al., 2011; Rosenfield, 1997; Wright et al., 2000). According to modified labeling theory (MLT) (Link, 1987; Link et al., 1989, 1997; Link et al., 1991), these negative consequences develop through three interrelated processes. First, when individuals are diagnosed with a psychiatric disorder, negative societal conceptions associated with the new label (e.g., incompetent, dangerous) become personally relevant and damage feelings of self-worth. Second, a psychiatric diagnosis that is publically known increases patients’ vulnerability to negative evaluation and rejection. Finally, patients whose self-concepts have been damaged through the first two processes adopt behaviors aimed at warding off rejection: concealing treatment history, withdrawing from social interactions, and educating others about mental illness. But, rather than helping ⇑ Corresponding author. E-mail address: [email protected] (A. Kroska). http://dx.doi.org/10.1016/j.ssresearch.2015.06.021 0049-089X/Ó 2015 Elsevier Inc. All rights reserved.

A. Kroska et al. / Social Science Research 54 (2015) 68–79

69

patients, these defensive behaviors harm them by, for example, constricting support networks and limiting employment opportunities. Thus, according to MLT, diagnostic labels damage patients by producing a negative self-concept, increasing others’ criticism and rejection, and triggering defensive behaviors that reduce social and employment success. Although the first and third MLT processes have been investigated in recent studies (Kroska and Harkness, 2006, 2008, 2011; Markowitz, 1998; Markowitz et al., 2011; Rosenfield, 1997; Wright et al., 2000), the second step—the increase in negative evaluation and rejection—has received little attention in recent years, particularly with studies that use a behavioral measure of rejection. Our study begins to address this gap by offering a contemporary examination of the causal link between psychiatric labels and negative evaluation and rejection. Building on other work in this area (Lucas and Phelan, 2012), we focus on competence-based evaluations and rejections within dyads. We operationalize evaluations with perceptual indicators (ratings of a person’s status and power) and rejection with behavioral indicators (rejecting a person’s suggestions when working together to solve a problem). We go beyond the existing work by exploring the way that shared gender and shared education affect these outcomes.

1.1. A psychiatric label We examine the effects of psychiatric labels on competence-based outcomes within the scope conditions of status characteristics theory (SCT) (Berger et al., 1966, 1972; Berger et al., 1977; for an overview, see Berger and Webster, 2006). According to SCT, when individuals in a group work together on a valued task, the diffuse status characteristics that differentiate them shape their expectations about how they and others will perform on the task. Diffuse status characteristics are culturally defined characteristics (e.g., gender) whose states (e.g., female, male) are given different degrees of esteem in the dominant culture. Those in the status advantaged group are expected to perform better than those in the status disadvantaged group, and individuals in both groups are expected to adopt these expectations. Consequently, the expectations function as self-fulfilling prophecies: individuals in the status-disadvantaged group, sensing that they have less to contribute than those in the advantaged group, participate less frequently and defer to those in the advantaged group more frequently, while those in the status-advantaged group, sensing that they have more to offer, participate more frequently and defer less readily. The differential expectations also shape the way performances are evaluated, with individuals perceiving the performance of the status-advantaged group as more valuable than that of the status-disadvantaged group, even when the performances are identical. Individuals occupying psychiatric identities are viewed as less intelligent, less wise, and less powerful than those who do not occupy such identities (Francis and Heise, 2006; Nunnally, 1961; Olmsted and Durham, 1976), suggesting that a psychiatric diagnosis may function as a diffuse status characteristic. Yet, the empirical work examining these processes in task-related encounters suggests that mental illness may function as a status characteristic only for men. This line of research began in the 1970s with experiments conducted by Farina and his colleagues. In one series of experiments, participants were asked to make hiring recommendations after interviewing job applicants, decisions that were likely based on perceptions of the applicants’ competence (Farina et al., 1973, 1978; Farina and Hagelauer, 1975). The male participants recommended hiring applicants described as psychiatric patients less frequently than applicants who were not described this way, a pattern that held for both male and female applicants, whereas female participants recommended hiring the psychiatric patients just as frequently as the non-patients, a pattern that also held for both male and female applicants. A related pattern emerged in an experiment wherein participants were asked to use electric shocks to help train another person of the same sex to learn a pattern of button presses (Farina et al., 1976). (Shocks were not actually administered.) Men in this study used lengthier shocks when they thought they were training a person with a psychiatric disorder, suggesting that they thought the patients needed stronger punishments to learn the patterns. Women, by contrast, used shorter shocks when training a person they thought had a psychiatric disorder, suggesting that they saw the psychiatric patients as more vulnerable or as able to learn with less punishment. Thus, experiments that examine competence-based assessments and rejections in task-oriented situations suggest that men treat psychiatric patients as less competent than non-patients, while women do not.1 Yet, the studies from the 1970s are not fully in compliance with the scope conditions of SCT. Although they examine rejection in a task-oriented setting, in contrast to the SCT scope conditions, the participant and the confederate in these studies were not working together on a task. Recently, however, Lucas and Phelan (2012) examined competence-based rejection in a way that complies with SCT scope conditions: within the confines of a collectively oriented task group. In line with the studies from the 1970s, they find that men are more likely to resist influence from teammates with a history of hospitalization for ‘‘psychological problems’’ (p = .040, one-tailed test), whereas women’s resistance to influence is unaffected by a teammate’s psychiatric hospitalization history (p = .106, one-tailed test).2 Thus, their findings suggest even more clearly than the studies in the 1970s that mental illness functions as a status characteristic for men but not for women. 1 Sibicky and Dovidio (1986) did find that both men and women judged cross-sex conversation partners who were described as mental health clients as less competent than cross-sex partners who were not described this way (same-sex pairs were not examined). But, this encounter was purely social, suggesting that SCT scope conditions were not fulfilled. Moreover, the exclusive use of cross-sex pairs may have masked women’s higher tolerance, given that male psychiatric patients are evaluated more negatively than female psychiatric patients. 2 Lucas and Phelan (2012) did not report gender-specific p-values in their article, but Jeff Lucas gave them to us in an email on December 11, 2013.

70

A. Kroska et al. / Social Science Research 54 (2015) 68–79

This gender difference is likely rooted in gendered behavior expectations: females are expected to behave in more communal, helpful, and less domineering ways than are men, while men are expected to behave in more assertive and forceful ways (Eagly and Karau, 2002). The prescriptive beliefs are communicated through socialization processes throughout the life course: females receive greater reinforcement for tolerant, communal, and cooperative behavior and stronger social reprisals for assertive and agentic behavior, while males receive greater reinforcement for assertive and aggressive behavior and stronger sanctions for communal behavior (e.g., Eagly and Karau, 2002; Ferguson, 2000; McGuffey and Rich, 1999; Messner, 2009; Pascoe, 2007; Rudman, 1998; Rudman et al., 2012). And these differential behavior expectations may be especially likely to produce gender-differentiated behavior when interacting with a person perceived as vulnerable or weak, such as a psychiatric patient. In line with these expectations, women and adolescent females consistently report valuing altruism and compassion more and materialism less than their male counterparts (Beutel and Johnson, 2004; Beutel and Marini, 1995), and young women place greater importance than young men on the altruistic, intrinsic, and social aspects of a job (Johnson et al., 2012; Marini et al., 1996). Women are also less likely than men to behave dominantly in a high status position when their higher status has not been widely legitimated (Ridgeway et al., 2009). In addition, women see themselves (e.g., myself as I really am) as higher in goodness and lower in power than men see themselves (Kroska, 2002), and both women and men rate female identities higher in goodness and lower in power than the male counterparts (Francis and Heise, 2006; Langford and MacKinnon, 2000). Yet, questions remain about these processes. Participants in Lucas and Phelan’s study were not told the gender of their teammate, so the patient’s gender was ambiguous. This ambiguity is a limitation because female psychiatric patients have historically been treated with greater tolerance than male psychiatric patients (see reviews in Farina, 1981, 1998). We begin to address this issue in our study by giving participants information on their teammate’s gender. We only examine same-sex dyads, so we cannot separate participant gender from patient gender. But, we can determine if gender differences exist in same-sex pairs when gender is made clear. Our design also specifies the type of psychiatric hospitalization (depression or schizophrenia), adding further clarity to the mental health status of the teammate. We also include a condition wherein the teammate is hospitalized for a non-psychiatric illness (food poisoning), allowing us to explore these processes for a non-psychiatric aliment. In addition, we explore the way the dyad’s shared educational attainment affects these patterns. Drawing on modified labeling theory and empirical work in this area, we advance the following hypotheses for competence-based rejections (operationalized with resisting influence when working together on a task) and competence-based evaluations (operationalized with ratings of status and power): H1a. A same-sex teammate’s history of psychiatric hospitalization will increase men’s resistance to the teammate’s influence but will not affect women’s.

H1b. A same-sex teammate’s history of psychiatric hospitalization will reduce men’s perceptions of the teammate’s competence but will not affect women’s. 1.2. Education According to status characteristics theory, education should not influence teammate perceptions and resistance behavior when individuals are working with others with the same education. But, we consider an aspect of education that is unrelated to status: its liberalizing effect on gendered behaviors. Education increases gender egalitarian beliefs (Cunningham et al., 2005; Davis and Robinson, 1991; Fan and Marini, 2000; Kroska and Elman, 2009), an effect that occurs even within the short window of time that students are in college (Astin, 1993; Bryant, 2003; Funk and Willits, 1987; Lottes and Kuriloff, 1994; Pascarella and Terenzini, 1991). Education also fosters more egalitarian behavior, including less domineering and more cooperative behavior in men and independent and assertive behavior in women (Baxter et al., 2008; Ellison and Anderson, 2001; Sanchez and Thomson, 1997). Therefore, in the context of problem-solving groups, we propose that education may moderate the effect of gender by increasing women’s resistance to influence and negative evaluations and reducing men’s. The teammate’s educational level is matched to participant’s, so education in our study reflects the education of both the participant and the teammate. H2a. Education will moderate the effect of gender on resistance to influence from a same-sex teammate by reducing men’s resistance and increasing women’s.

H2b. Education will moderate the effect of gender on perceptions of a same-sex teammate by increasing men’s perceptions of the teammate’s competence and decreasing women’s perceptions of the teammate’s competence. Yet, because we anticipate that mental illness will function as a status characteristic for men, we expect the psychiatric status of the teammate to moderate the effect of education among men. Specifically, we expect the negative effect of education on men’s resistance to influence (H2a) and perceptions of their teammate’s competence (H2b) to be stronger when the

A. Kroska et al. / Social Science Research 54 (2015) 68–79

71

teammate is equal in status (not a psychiatric patient) rather than lower status (a psychiatric patient). Because we do not expect mental illness to function as a status characteristic among women, we do not expect the teammate’s psychiatric history to moderate the effect of education among women. Thus, we hypothesize a three-way effect with the teammate’s psychiatric status moderating the two-way interaction between education and gender in both the resistance to influence models and perceptions of competence models: H3a. A same-sex teammate’s history of psychiatric hospitalization will weaken the effect of education on resistance to influence among men but have no effect on that relationship among women.

H3b. A same-sex teammate’s history of psychiatric hospitalization will weaken the effect of education on perceptions of a teammate’s competence among men but have no effect on that relationship among women.

2. Methods 2.1. Sample We collected data from 199 undergraduates at a public university in the south during the 2010–11 academic year. In the debriefing, 16 of these participants reported a very clear and early suspicion that there was no teammate and/or that the joint task was not real, and three elected to have their data destroyed, a standard option in the debriefing form, leaving 180 non-suspicious participants (55 males and 125 females) who were willing to have their data retained. Motivation for success on the joint task, termed ‘‘the meaning insight task,’’ is an important criterion for inclusion in SCT studies, so we excluded the cases with the ten lowest responses (six percent) to the question ‘‘How important was it for you to get the right answer on the meaning insight task?’’ These ten cases gave responses that ranged from 0 to 27 on a 0–100 scale. With this exclusion, we have 170 retainable cases (49 males and 121 females), with an overall of exclusion of 13.3% (26/196) among cases willing to be analyzed, an exclusion rate that is a bit below the average rate (14.53%) among SCT studies that report doing exclusions according to Dippong’s (2012) meta-analysis. Rates of exclusion by condition are 18.6% in the schizophrenia condition, 10.9% in the depression condition, 17.2% in the food poisoning condition, and 6.1% in the non-patient condition. The rate is 9.7% among females and 21.0% among males. The difference in rates of exclusion between the schizophrenia and nonpatient conditions (p = .08; two-tailed test) and between the food poisoning and nonpatient conditions (p = .09; two-tailed test) are close to significance, and the difference by gender is significant (p = .03; two-tailed test). 2.2. Teammate’s hospitalization history We use a self-reported history of psychiatric hospitalization to operationalize a psychiatric label. At the beginning of the computerized instructions, participants learned that they would be working with a teammate on 25 ‘‘meaning insight tasks.’’ The instructions then asked them to fill out an electronic information sheet that would be exchanged with the teammate. The instructions explained that ‘‘the educational, employment, and demographic information you exchange will be similar to the information you might obtain from co-workers at a job.’’ The instructions also asked participants to, ‘‘Please answer the following questions about yourself carefully and accurately.’’ The form asked participants their gender, age, year in college, years of work experience, type of work experience, and whether they had had to take a leave of absence from school or work, and, if so, the reason. The teammate’s responses to the last two questions served as the manipulation of the teammate’s hospitalization history, and these responses were randomly assigned by the computer program. In the non-patient condition, the teammate response to the leave of absence question was simply ‘‘No.’’ In the depression, schizophrenia, and food poisoning conditions, the answer was ‘‘Yes,’’ and the answer to the reason question was ‘‘Last year I was hospitalized for depression/schizophrenia/food poisoning, so I took a little time off.’’ Table 1 shows the descriptive statistics for this and the other variables in the analyses. The teammate’s responses were matched with the participant’s on all the other questions so as not to introduce any other status differences, and we used broad response categories for all these questions except year in college so that the matching responses did not arouse suspicion. After participants were shown the teammate’s information, the instructions asked them to write the information down on a ‘‘Partner Information Sheet’’ beside the computer, a task that ensured that participants in the psychiatric conditions saw the hospitalization information. The manipulations created four conditions: non-patient teammate; teammate hospitalized for depression; teammate hospitalized for schizophrenia; and teammate hospitalized for food poisoning. Because the teammate was paired with the participant on both gender and education, those attributes also vary across conditions and are, therefore, controlled in the models as well. 2.3. Meaning insight task After exchanging information with the teammate, participants learned more about the meaning insight tasks (MIT), a standard task used for investigating status-organizing processes (Berger, 2014). Participants learned that on each of

72

A. Kroska et al. / Social Science Research 54 (2015) 68–79

Table 1 Descriptive statistics for variables in analyses (N = 170). Men (N = 49)

Dependent variables Percentage of staysa Perception of teammate’s status Perception of teammate’s potency Independent variables Mentally ill teammateb Educationc

Women (N = 121)

Mean

SD

Min

Max

Mean

SD

Min

Max

.44 61.00 .42

.15 12.60 1.20

.03 40.50 4.3

.73 96.33 3.0

.42 62.80 .24

.15 13.11 1.38

0 34.50 4.3

.88 100 3.0

.37 .61

.49 .84

.48 .52

.50 .80

0 0

1 3

0 0

1 3

a Weighted inversely by the popularity of the first selection and more heavily on trials with a low stay rate; also used as an independent variable in one analysis. b 0 = teammate is non-patient or food poisoning patient. c Attribute of participant and teammate.

the 25 meaning insight tasks, the two teammates would be presented with two words from a reconstructed language (e.g., yut-ken or yan-tek) and one English word (e.g., sharp) and that their task was to determine which of the two words was most likely to be related to the English word. Through an example trial, the teammates learned that they would provide an initial answer that was shared and that each teammate would then privately enter his or her final answer. In reality, there were no correct answers, and the partner was computerized and programmed to give an initial answer that differed from the participant’s on 20 of the trials (all but trials 1, 6, 13, 17, and 22). Participants were told that the teammates’ final choices on each trial would be combined and that the team with the highest number of correct answers that semester would split a $100 bonus. This joint reward was designed to create a valued outcome and to motivate participants to work with the teammate to find the correct answer, contributing to the fulfillment of two SCT scope conditions (valued outcome and collective orientation). After the 25 trials, participants completed a post-experimental questionnaire. 2.4. Other independent variables The teammate’s gender and education were matched to the participant’s. Education ranges from 0 (freshman) to 3 (senior). 2.5. Dependent variables Resistance to influence is operationalized as the percentage of the 20 disagreement trials in which participants stay with their initial choice for their final choice in the MIT. Although the MIT is designed to give participants two equally plausible word options, the initial selections differed from a 50–50 divide for 14 of the 20 disagreement trials (p < .05, two-tailed tests), suggesting that the two options were not perceived as equally plausible on all trials. Hence, we created a stay score that weights each stay inversely by the popularity of the initial selection and that gives a greater weight to stays on trials with a low stay rate. First, we multiplied the participants’ stay score (0 or 1) by the absolute value of the difference between their initial answer (coded as 1 or 2) and the average initial answer (1.5 if the selections were equally divided between the two choices). For example, on trial 2, 47.06% of the participants selected the option on the left (pa-le) (coded as 1), and 52.94% selected the option on the right (se-weh) (coded as 2), so the average score for trial 2 was 1.5294. Thus, participants who selected the left option (the less popular option) and stayed with it were given a higher stay score (.5294) than those who selected the right option and stayed with it (.4706). Then, we multiplied that weighted stay score by a second weighted stay, which we created by dividing the stay score (0 or 1) by the percent who stayed for that trial. We created both these weights using only the 170 cases in this sample. The distribution of the variable is normal (chi square for the joint test of skewness and kurtosis = 2.28, p = .32). We also operationalized the percentage of stays in three other ways.3 The version we present has the distribution that is closest to normality and provides the greatest explained variance, but the differences across models are very small. The results with the other versions are available from the first author on request. Perception of teammate’s status is a summed average of six items measured on 101-point sliders: not respected/respected, low status/high status, follower/leader, incompetent/competent, unknowledgeable/knowledgeable, and incapable/capable (alpha reliability = .86). The direction and order of the items were randomized across participants. The instructions introducing these measures emphasized that the partner would not see the ratings. Although respect items (respect, status, leader) are sometimes separated from competence items (competence, knowledgeable, capable) to create two constructs (e.g., Ridgeway et al., 2009), our analyses show that these six items tap a single construct. The eigenvalue for the first factor using principal-factor factor analysis is 3.10 but only .21 for the second. In addition, the alpha reliability scores for the three-item 3 The three ways are the traditional, unweighted approach (percentage of 20 disagreement trials in which participants stay with their initial answer for their final answer) and versions that use only one of the two weights included in the version we use. The bivariate correlations among all pairs of the four versions is .95 or higher.

73

A. Kroska et al. / Social Science Research 54 (2015) 68–79

composites (.85 for competence and .59 for respect) are lower than the alpha for the six-item composite. Therefore, we collapsed all six items into a single composite. Perception of teammate’s power reflects the participants’ rating of ‘‘my partner’’ on a single 9-point semantic differential scale anchored with ‘‘powerful, big’’ and ‘‘powerless, small.’’ The middle point of the scale was marked ‘‘neutral’’; the points between the midpoint and the endpoints were marked ‘‘slightly,’’ ‘‘quite,’’ ‘‘extremely,’’ and ‘‘infinitely’’ and were coded as 4.3, 3, 2, 1, 0, 1, 2, 3, 4.3. The direction of the adjective pairs was randomized across participants. The instructions introducing the scale emphasized again that the ratings would not be shared with the partner.

3. Results 3.1. Resistance to influence We first examined the percentage of stay results separately by condition and found that the results in the two psychiatric conditions and the two non-psychiatric conditions were sufficiently similar to warrant collapsing the conditions to create one dichotomous variable (mentally ill teammate vs. not). The similarity between the control and the food poisoning results suggests that a food poisoning diagnosis does not function as a status characteristic. Table 2 displays the OLS regressions of the weighted stay rate on condition and controls. According to H1a, a same-sex teammate’s history of psychiatric hospitalization should increase men’s resistance to influence but have no effect on women’s. Model 1 fails to support that hypothesis, because the positive effect of a teammate’s psychiatric hospitalization on men’s resistance does not reach significance (b = .071, p = .061; one-tailed test).4 But, Model 2, which controls for the gendered effects of education, provides support. As shown, a same-sex teammate’s psychiatric hospitalization increases men’s resistance (b = .108, p = .012), is unrelated to women’s resistance (b = .024, p = .185), and this difference is significant (b = .133, p = .008). With H2a, we predicted that education would reduce men’s resistance to influence but increase women’s. Model 2 provides partial support for this hypothesis as well, showing that education reduces men’s resistance (b = .065, p = .009), an effect that is significantly different from the effect among women (b = .083, p = .006). However, contrary to predictions, education does not increase women’s resistance (b = .018, p = .148). Model 3 introduces the interaction between education and condition, and Model 4 adds the three-way term needed to test Hypothesis 3a. The results in Model 4 support H3a. As predicted, the two-way interaction between education and

Table 2 OLS regressions of percentage of stays on condition and controls (N = 170). Percentage of staysa

Model

1 b

Mentally ill teammate c

Female

Educationc Female  mentally ill teammate

2

3 *

4 *

.071 (.046) .018 (.033) .005 (.015) .095* (.054)

.108 (.047) .017 (.036) .065** (.028) .133** (.055) .083** (.032)

.099 (.051) .018 (.036) .076* (.034) .132** (.055) .086** (.033) .016 (.030)

.419 (.028) .022 .002

.443 (.029) .060 .031

.447 (.030) .062 .027

Female  education Education  mentally ill teammate Female  education  mentally ill teammate Intercept 2

R Adjusted R2

.040 (.058) .053 (.040) .145** (.049) .056 (.067) .171*** (.054) .116* (.059) .134* (.068) .473 (.033) .084 .044

Note: Unstandardized coefficients; standard errors are in parentheses. * p < .05. ** p < .01. *** p < .001 (one-tailed tests). a Weighted inversely by the popularity of the first selection and more heavily on trials with a low stay rate. This is the measure of resistance. b 0 = teammate is non-patient or food poisoning patient. c Attribute of participant and teammate. 4 Most of our predictions are directional, so following reviewer recommendations, we use a one-tailed test to determine the p-value for the effects in this study. In the non-directional predictions, we expect no effects (e.g., the effect of the teammate’s psychiatric status among women), and the one-tailed test provides a more stringent test of that expectation.

A. Kroska et al. / Social Science Research 54 (2015) 68–79

.4 .3 .2 0

.1

Weighted Stay Rate

.5

.6

74

Freshmen

Sophomores

Juniors

Seniors

Education Females, MI teammate Females, Non-MI teammate

Males, MI teammate Males, Non-MI teammate

Fig. 1. Weighted stay rate by MI teammate, gender, and education.

condition is positive and significant among men (b = .116, p = .025) but non-significant among women (b = –.017, p = .305), and the difference in these slopes is significant, as shown by the three-way term (b = .134, p = .025). The results show that education reduces men’s resistance when their partner has no psychiatric history (b = .145, p = .002) but has no significant effect when the partner has a psychiatric history (b = .028, p = .198). Thus, as we anticipated, the liberalizing effect of education on task behavior is strongest among men working with a status equal; in fact, when working with a lower status partner, the effect is weakened to non-significance. Fig. 1 provides a plot of the predicted values generated from Model 4. As shown, education reduces men’s resistance, but the decline is steeper for those who are working with a status equal.5 Table 3 OLS regressions of perceptions of teammate’s status and power on condition and controls (N = 170). Model

Mentally ill teammatea Femaleb Educationb Female  mentally ill teammate

Status

Power

1

2

1.20 (3.93) 2.07 (2.85) 1.77 (1.26) .51 (4.58)

.79 (2.05) 2.07 (2.75) 2.03 (2.28)

Female  education

3 .54 (.40) .35 (.29) .03 (.13) .49 (.47)

.34 (2.72)

Proportion mediated by % of stays 4 .85* (.42) .06 (.31) .54* (.24) .80* (.48) .70** (.29)

Percentage of staysc Intercept 2

R Adjusted R2

59.47 (2.38) .018 .005

59.47 (2.36) .018 .005

.60 (.24) .015 .009

.40 (.25) .050 .021

p-value of Sobel test (one-tailed)

5 .69 (.42) .08 (.31) .45* (.24) .61 (.48) .58* (.29) 1.47* (.68) 1.05 (.39) .077 .043

.19

.058

.46

.316

.18

.055

.24

.053

.17*

.049

Note: Unstandardized coefficients; standard errors are in parentheses. * p < .05. ** p < .01 (one-tailed tests). a 0 = teammate is non-patient or food poisoning patient. b Attribute of participant and teammate. c Weighted inversely by the popularity of the first selection and more heavily on trials with a low stay rate. This is the measure of resistance.

5 Given the small number of juniors and seniors in our sample (10 juniors and 8 seniors), we considered collapsing them into one education category. When we do this, the results are highly similar. The R2 is slightly higher in Models 1 and 2 but slightly lower in Models 3 and 4, and some terms become slightly more significant (e.g., the Model 2 education slope for men who are not working with a mentally ill teammate is b = .085, p = .006 rather than b = .065, p = .009), while others become slightly less significant (e.g., the Model 4 education  mentally ill teammate term is b = .113, p = .046 rather than b = .116, p = .025, and the three-way term is b = .140, p = .037 rather than b = .134, p = .025). In the end, we kept the four category version of education so that we could retain the integrity of the measure (i.e., not merge participants with different education levels), retain more variability in the variable, and more fully test our hypothesis that each additional year of education affects task behavior in a gendered way. However, the results using the collapsed version of education are available from the first author on request.

A. Kroska et al. / Social Science Research 54 (2015) 68–79

75

3.2. Perceptions of status and power Table 3 displays the OLS regressions of partner status and partner power ratings on condition and controls.6 According to H1b, a same-sex partner’s psychiatric hospitalization should reduce men’s perceptions of the partner’s status and power but have no effect on those perceptions among women. As Model 1 shows, psychiatric hospitalization has no effect on perceptions of the partner’s status for women or men, a pattern that fails to fully support the hypothesis. But, we find more complete support for this hypothesis with perceptions of power. Although a teammate’s history of psychiatric hospitalization is not significantly related to men’s impressions of his power in Model 3 (b = .54, p = .091), in Model 4, which controls for the gendered effects of education, the effect becomes significant. As shown, men perceive psychiatric patient partners as less powerful than non-psychiatric partners (b = .85, p = .021), while women’s power ratings are unaffected by their partner’s psychiatric history (b = .05, p = .422), a slope difference that is significant (b = .80, p = .048). According to H2b, education should moderate the effect of gender on power and status assessments. As shown in Model 2, we find no support for this hypothesis for status ratings. But, as shown in Model 4, we find partial support with power ratings. As predicted, education increases men’s perceptions of their partner’s power (b = .54, p = .013), and the effect is significantly different from the effect among women (b = .70, p = .007). But, contrary to predictions, the negative effect of education among women does not reach significance (b = .16, p = .142). According to H3b, the teammate’s psychiatric status should moderate the positive effect of education on men’s status and power ratings and have no effect on these relationships among women. But, contrary to predictions, men’s teammate’s psychiatric status does not alter the (non-significant) effect of education on partner status ratings (not shown) nor does it weaken the positive effect of education on partner power ratings (not shown). And, although women’s teammate’s psychiatric status also does not alter the effect of education, as predicted (not shown), the non-significance of the education effect makes the result moot. Thus, we fail to find support for H3b. 3.2.1. Stay behavior as mediator The partner perceptions were reported after the meaning insight task, so it is possible that stay behavior during the task mediated the significant relationships identified in Model 4. We examine that possibility by controlling for the percentage of stays in Model 5 and report in the next two columns the proportion of each coefficient mediated by the percentage of stays and the significance of the mediation using the Sobel test. One of the mediation effects reaches significance, and three others are just outside the cut-off for significance, suggesting that the positive effect of education and the negative effect of a mentally ill teammate on men’s partner power ratings may be partially mediated by stay behavior. More specifically, education reduces men’s resistance (b = .065, p = .009; see Model 2 in Table 2), and that reduced resistance then elevates impressions of the partner’s power, while a mentally ill teammate increases men’s resistance (b = .108, p = .012; see Model 2 in Table 2), and that elevated resistance then reduces impressions of the partner’s power. However, given the non-significant p-values for most of the mediation effects, the results simply raise the possibility of mediation; further tests are needed to corroborate the findings. 4. Discussion According to modified labeling theory, the negative consequences of psychiatric labeling develop through three interrelated processes that begin at diagnosis. First, the personal relevance of negative cultural conceptions regarding mental illness creates feelings of demoralization. Next, the publically known mental illness increases others’ tendency to criticize and reject the patient. Then, as the first two processes unfold, the third process begins: patients start using coping behaviors—secrecy, withdrawal, and/or education—to prevent rejection, behaviors that, ironically, are hypothesized to ultimately harm them by reducing support networks and employment opportunities. The first and third processes have been examined in several recent studies (e.g., Kroska and Harkness, 2011; Markowitz et al., 2011), but the second process has received less attention in recent decades, particularly with studies that use a behavioral measure of rejection. We examined that process, focusing specifically on competence-based assessments (status and power ratings) and rejections (ignoring a teammate’s suggestions) within dyads comprised of same-sex individuals with equivalent education who were working together to solve a series of problems. We also explored the effects of gender and education on these processes. 4.1. Psychiatric hospitalization 4.1.1. Resistance to influence We found, as predicted, that men ignored suggestions from teammates with a history of psychiatric hospitalization more frequently than they ignored suggestions from teammates with no such history. The computerized partners interacted with the participant in the same way across conditions, so this finding suggests that a psychiatric label alone is enough to prompt men’s resistance to influence. Social interactions in laboratory experiments are unusual, because participants are highly 6 The results for status differed across the four conditions to a greater extent and in different ways than in the behavioral results, so we tried coding the conditions in various ways, including leaving depression, schizophrenia, and food poisoning as separate dummy variables. However, the results remained largely non-significant across the different coding schemes, so we stayed with the coding we used for the behavioral analyses. Those results are available from the first author on request.

76

A. Kroska et al. / Social Science Research 54 (2015) 68–79

aware that they are being observed and are, most likely, self-conscious about their behavior. Yet, those features of the situation make this finding even more striking: despite the oversight and likely self-awareness, the men paired with a psychiatric patient dismissed their partners’ suggestions at a higher rate than the men paired with a non-patient or a food poisoned patient. This finding suggests that a psychiatric diagnosis functions as a status characteristic in all male groups and that the second step in the modified labeling theory process—rejection after a psychiatric diagnosis becomes public—applies to collectively oriented, male task groups. Women, however, were not more dismissive of their teammate’s suggestions when the teammate disclosed a history of psychiatric hospitalization. As predicted, women were just as likely to defer to a same-sex partner with a history of psychiatric hospitalization as a same-sex partner with no such history. Thus, despite declines in gender differences in several realms, including education (Buchmann and DiPrete, 2006), labor force participation (Cotter et al., 2011), and domestic work time (Sayer, 2005; Sayer et al., 2004), women and men may still behave differently when interacting with a vulnerable person. These differential behavior patterns are likely rooted in prescriptive gender stereotypes and the concomitant socialization processes and expectational pressures that promote nicer, less domineering, and more tolerant behavior among females (e.g., Eagly and Karau, 2002; Rudman et al., 2012). Yet, additional studies that include both same-sex and cross-sex pairs are needed to determine how much of this gender difference is due to the gender of the participant and how much is due to the gender of the patient. Indeed, women’s higher tolerance could be a function of the gender of the patient with whom they were working, given the greater tolerance typically shown to female psychiatric patients (Farina, 1981). But, the correspondence between these results and studies that included cross-sex pairs (Farina et al., 1973, 1978; Farina and Hagelauer, 1975) or left the sex of the patient ambiguous (Lucas and Phelan, 2012) suggests that at least some of the effect is due to the gender of the participant. Other studies suggest that women and men treat psychiatric patients similarly when the setting is purely social (Kroska et al., 2014; Lucas and Phelan, 2012; Sibicky and Dovidio, 1986), so gender difference in the treatment of psychiatric patients may only emerge in the confines of collectively oriented task groups. But, future studies that include both same-sex and cross-sex pairs in both task groups and purely social settings are needed to fully explore these differences. 4.1.2. Status and power We also examined two non-behavioral outcomes—perceptions of the partner’s status and power—and found, as predicted, that psychiatric hospitalization reduced men’s perceptions of their partner’s power. However, this effect may have been partially mediated by stay behavior, suggesting that the act of ignoring the patient’s task suggestions contributed to the impression of him as weak. Future work exploring these processes could examine if and how rejection itself contributes to weak conceptions of individuals with psychiatric illnesses. Contrary to our predictions, however, a partner’s psychiatric illness did not affect men’s reported perceptions of their partner’s status. The contrast between the results for this explicit measure of status and the more subtle behavioral measure suggests that the perceptual results reflect men’s social desirability concerns. The perceptual measures may be gauging overt, or explicit, beliefs, which are more affected by social desirability concerns, while the more subtle behavioral measure may be tapping unconscious, or implicit, beliefs, which are less affected by those concerns (Dovidio et al., 2001; Greenwald and Banaji, 1995). Inconsistencies between attitudes and behavior and between explicit and implicit measures are a well-documented pattern in stigma research (Crocker et al., 1998; Kroska et al., 2014; Stier and Hinshaw, 2007) and in the SCT literature (Rashotte and Webster, 2005). In fact, these patterns are consistent with SCT assumptions, which hold that hierarchical task behaviors are likely rooted in implicit, or sub-conscious, status beliefs that are often inaccessible when measured explicitly (e.g., Berger et al., 1972, 1977; Ridgeway, 1997). The contrast between the perceptual status and power results may also be related to social desirability concerns. The endpoints for the power measure—‘‘powerful, big’’ and ‘‘powerless, small’’—are a bit more abstract and therefore less explicit than the endpoints for the items in the status composite (e.g., competent to incompetent, capable to incapable). In addition, the instructions above the power measure repeated an earlier statement about the ratings not being shared with the partner, so when participants provided the power ratings, they had read that instruction an additional time, a reiteration that may have increased their candor and/or the accessibility of their implicit beliefs. 4.2. Education Education increases gender egalitarian beliefs, a pattern that is evident even within the four year window that students are in college (e.g., Astin, 1993; Bryant, 2003; Lottes and Kuriloff, 1994; Pascarella and Terenzini, 1991). Education also fosters gender egalitarian behaviors, increasing gentleness and contributions to housework among men and assertiveness among women (e.g., Baxter et al., 2008; Ellison and Anderson, 2001). Given these patterns, we predicted that education would reduce gender-typical behavior (H2a) and evaluations (H2b) in the task group. But, because we expected mental illness to function as a status characteristic among men, we also predicted that these education effects would be weakened when men were paired with a teammate with a psychiatric history (H3a and H3b). We found support for H2a and H3a among men. As predicted, education reduced men’s overall resistance, but the psychiatric status of the partner moderated this effect. When the partner had no psychiatric history, education reduced resistance, but when the partner had a psychiatric history, the education effect weakened to non-significance. We also found some support for H2b among men, with education increasing men’s impressions of their partner’s power, although some of this effect may have been mediated

A. Kroska et al. / Social Science Research 54 (2015) 68–79

77

by stay behavior: education reduced men’s resistance to influence, which, in turn, elevated their perceptions of the partner’s power. Contrary to H3b, however, men’s partner’s psychiatric status did not alter the effect of education on partner power ratings. Also contrary to predictions, education had no effects among women: it failed to significantly heighten women’s resistance to influence, and it did not reduce the status and power they saw in their partner. Together the education results suggest that when interacting with status equals, a college education fosters deferential behavior, perhaps even humility, among men but has no significant effect on deference among women. We see several avenues for future work exploring this education effect. First, future studies should include a larger pool of upper-division students and more variation in educational attainment. The education effects that did not reach significance here—i.e., among women and among men working with psychiatric patients—might reach significance with a larger pool of more educated participants and with more variation on education. Future studies could also focus on identifying the mechanisms of education that create these changes. These mechanisms may be aspects of college curriculum, such as humanities or social science classes that examine gender inequality, and/or the experience of working collaboratively in classes or in campus organizations. Such findings will help clarify the link between education and task behavior while also identifying experiential tools that may promote assertiveness in women and communalism in men. Finally, future studies could explore these processes with other indicators of gender egalitarianism, such as gender attitudes or political values. 4.3. Generalizability to other populations College students differ in important ways from non-college students, and college students at public universities in the south differ from other types of college students, so future tests of these propositions on other populations—private universities, universities outside of the south, and non-university populations—will help determine the generalizability of our findings. Some analysts argue that the generally liberal values of college and university students make college students less likely than others to display prejudicial behaviors, a tendency that presents a challenge to experimentalists who study prejudice using college student samples (Henry, 2008; Sears, 1986, 2008). If these analysts are correct, our study and others like it do underestimate the extent to which individuals reject psychiatric patients’ task-related suggestions. Future replications that include non-student samples could explore this possibility and determine if college enrollment moderates these processes. Given that political values are the hypothesized predictor variable, future studies also could directly assess this possibility by measuring political and social attitudes prior to task group behavior to determine if these attitudes affect deference to psychiatric patients or stigmatized individuals more generally. 4.4. Implications for modified labeling theory Our findings have two implications for modified labeling theory. First, the findings provide further evidence that the second step in the modified labeling theory process—rejection after a psychiatric diagnosis becomes public—applies not just to purely social situations, as other studies have shown (e.g., Kroska et al., 2014; Lucas and Phelan, 2012), but also to collectively oriented, male task groups. Future studies exploring this process could examine the way that displays of symptoms affect these patterns. Second, our findings suggest that the act of rejecting a psychiatric patient’s suggestions may lead the rejecter to see the patient as weak, possibly setting in motion a self-perpetuating cycle. Future studies could more directly examine this possibility by determining if and how rejection itself contributes to the negative conceptions associated with mental illness. 4.5. Implications for status characteristics theory Our findings have two implications for status characteristics theory. First, our study provides further evidence that a psychiatric diagnosis functions as a status characteristic among men but does not among women. However, cross-sex dyads are needed to determine how much of the gender difference is due to participant gender and how much is due to patient gender. But the similarity of our findings with others that included cross-sex pairs (Farina, 1981) or left patient gender ambiguous (Lucas and Phelan, 2012) suggests that at least some of the gender difference is rooted in the way that men and women feel they are expected to behave around a vulnerable person, such as a psychiatric patient. Second, our findings suggest that men’s education affects their task group behavior. More specifically, education may weaken gendered behavior among men such that highly educated men accept suggestions from each other at a higher rate than do less educated men. Future studies could explore this hypothesis by examining a wider range of behaviors, such as verbal domination and non-verbal gestures, to determine if men’s education affects the full range of behaviors evident in task groups. Acknowledgments We thank Troy Smith for computer programming, Emily Pain for assistance with the pilot, and Will Kalkhoff, Rob Clark, Alison Bianchi, and members of the Social Psychology Workshop at Stanford University for very helpful feedback. This research was supported by two University of Oklahoma grants: an Ed Cline Faculty Development Award and a Faculty Enrichment Grant.

78

A. Kroska et al. / Social Science Research 54 (2015) 68–79

References Astin, A.W., 1993. What matters in college? Four Critical Years Revisited. Jossey-Bass, San Francisco. Baxter, J., Hewitt, B., Haynes, M., 2008. Life course transitions and housework: marriage, parenthood, and time on housework. J. Marriage Family 70, 259– 272. Berger, J., 2014. The standardized experimental situation in expectation states research: notes on history, uses, and special features. In: Webster, M., Jr., Sell, J. (Eds.), Laboratory Experiments in the Social Sciences, second ed. Elsevier, Boston, pp. 269–293. Berger, J., Cohen, B.P., Zelditch Jr., M., 1966. Status characteristics and expectation states. In: Berger, J., Zelditch, M., Anderson, B. (Eds.), Sociological Theories in Progress. Houghton Mifflin, Boston, pp. 29–46. Berger, J., Cohen, B.P., Zelditch Jr., M., 1972. Status characteristics and social interaction. Am. Sociol. Rev. 37, 241–255. Berger, J., Fiek, M.H., Norman, R.Z., Zelditch Jr., M., 1977. Status Characteristics and Social Interaction: An Expectation-State Approach. Elsevier, New York. Berger, J., Webster Jr., M., 2006. Expectations, status, and behavior. In: Burke, P.J. (Ed.), Contemporary Social Psychological Theories. Stanford University Press, Stanford, pp. 268–300. Beutel, A.M., Johnson, M.K., 2004. Gender and prosocial values during adolescence: a research note. Sociol. Quart. 45, 379–393. Beutel, A.M., Marini, M.M., 1995. Gender and values. Am. Sociol. Rev. 60, 436–448. Bryant, A.N., 2003. Changes in attitudes toward women’s roles: predicting gender-role traditionalism among college students. Sex Roles 48, 131–142. Buchmann, C., DiPrete, T.A., 2006. The growing female advantage in college completion: the role of family background and academic achievement. Am. Sociol. Rev. 71, 515–541. Cotter, D., Hermsen, J.M., Vanneman, R., 2011. The end of the gender revolution? Gender role attitudes from 1977 to 2008. Am. J. Sociol. 117, 259–289. Crocker, J., Major, B., Steele, C., 1998. Social stigma. In: Gilbert, D., Fiske, S.T., Lindzey, G. (Eds.), The Handbook of Social Psychology, fourth ed. McGraw Hill, Boston, pp. 504–553. Cunningham, M., Beutel, A.M., Barber, J.S., Thornton, A., 2005. Reciprocal relationships between attitudes about gender and social contexts during young adulthood. Soc. Sci. Res. 34, 862–892. Davis, N.J., Robinson, R.V., 1991. Men’s and women’s consciousness of gender inequality: Austria, West Germany, Great Britain, and the United States. Am. Sociol. Rev. 56, 72–84. Dippong, J., 2012. The effects of scope condition-based participant exclusion on experimental outcomes in expectation states research: a meta-analysis. Soc. Sci. Res. 41, 359–371. Dovidio, J.F., Kawakami, K., Beach, K., 2001. Implicit and explicit attitudes: examination of the relationship between measures of intergroup bias. In: Brown, R., Gaertner, S.L. (Eds.), Blackwell Handbook of Social Psychology, Intergroup Relations, vol. 4. Blackwell, Oxford, UK, pp. 175–197. Eagly, A.H., Karau, S.J., 2002. Role congruity theory of prejudice toward female leaders. Psychol. Rev. 109, 573–598. Ellison, C.G., Anderson, K.L., 2001. Religious involvement and domestic violence among U.S. couples. J. Sci. Study Religion 40, 269–286. Fan, P., Marini, M.M., 2000. Influences on gender-role attitudes during the transition to adulthood. Soc. Sci. Res. 29, 258–283. Farina, A., 1981. Are women nicer people than men? Sex and the stigma of mental disorders. Clin. Psychol. Rev. 1, 223–243. Farina, A., 1998. Stigma. In: Mueser, K.T., Tarrier, N. (Eds.), Handbook of Social Functioning in Schizophrenia. Allyn and Bacon, Boston, pp. 247–279. Farina, A., Felner, R.D., Boudreau, L.A., 1973. Reaction of workers to male and female mental patient job applicants. J. Consult. Clin. Psychol. 41, 363–372. Farina, A., Hagelauer, H.D., 1975. Sex and mental illness: the generosity of females. J. Consult. Clin. Psychol. 43, 122. Farina, A., Murray, P.J., Groh, T., 1978. Sex and worker acceptance of a former mental patient. J. Consult. Clin. Psychol. 46, 887–891. Farina, A., Thaw, J., Felner, R.D., Hust, B.E., 1976. Some interpersonal consequences of being mentally ill or mentally retarded. Am. J. Mental Deficiency 80, 414–422. Ferguson, A.A., 2000. Bad Boys: Public Schools in the Making of Black Masculinity. University of Michigan Press, Ann Arbor. Francis, C., Heise, D.R., 2006. Mean affective ratings of 1500 concepts by Indiana University undergraduates in 2002-3. Distributed at affect control theory website, Program interact . Funk, R.B., Willits, F.K., 1987. College attendance and attitude change: a panel study. Sociol. Educ. 60, 224–231. Greenwald, A.G., Banaji, M.R., 1995. Implicit social cognition: attitudes, self-esteem, and stereotypes. Psychol. Rev. 102, 4–27. Henry, P.J., 2008. College sophomores in the laboratory redux: influences of a narrow data base on social psychology’s view of the nature of prejudice. Psychol. Inq. 19, 49–71. Johnson, M.K., Sage, R.A., Mortimer, J.T., 2012. Work values, early career difficulties, and the U.S. economic recession. Soc. Psychol. Quart. 75, 242–267. Kroska, A., 2002. Does gender ideology matter? Examining the relationship between gender ideology and self- and partner-meanings. Soc. Psychol. Quart. 65, 248–265. Kroska, A., Elman, C., 2009. Change in attitudes about employed mothers: exposure, interests, and gender ideology discrepancies. Soc. Sci. Res. 38, 366–382. Kroska, A., Harkness, S.K., 2006. Stigma sentiments and self-meanings: exploring the modified labeling theory of mental illness. Soc. Psychol. Quart. 69, 325– 348. Kroska, A., Harkness, S.K., 2008. Exploring the role of diagnosis in the modified labeling theory of mental illness. Soc. Psychol. Quart. 71, 193–208. Kroska, A., Harkness, S.K., 2011. Coping with the stigma of mental illness: empirically-grounded hypotheses from computer simulations. Soc. Forces 89, 1315–1339. Kroska, A., Harkness, S.K., Thomas, L.S., Brown, R.P., 2014. Illness labels and social distance. Soc. Mental Health 4, 215–234. Langford, T., MacKinnon, N.J., 2000. The affective basis for the gendering of traits: comparing the United States and Canada. Soc. Psychol. Quart. 63, 34–48. Link, B.G., 1982. Mental patient status, work, and income: an examination of the effects of a psychiatric label. Am. Sociol. Rev. 47, 202–215. Link, B.G., 1987. Understanding labeling effects in the area of mental disorders: an assessment of the effects of expectations of rejection. Am. Sociol. Rev. 52, 96–112. Link, B.G., Cullen, F.T., Struening, E., Shrout, P.E., Dohrenwend, B.P., 1989. A modified labeling theory approach to mental disorders: an empirical assessment. Am. Sociol. Rev. 54, 400–423. Link, B.G., Mirotznik, J., Cullen, F.T., 1991. The effectiveness of stigma coping orientations: can negative consequences of mental illness labeling be avoided? J. Health Soc. Behav. 32, 302–320. Link, B.G., Struening, E.L., Rahav, M., Phelan, J.C., Nuttbrock, L., 1997. On stigma and its consequences: evidence from a longitudinal study of men with dual diagnoses of mental illness and substance abuse. J. Health Soc. Behav. 38, 177–190. Lottes, I.L., Kuriloff, P.J., 1994. The impact of college experience on political and social attitudes. Sex Roles 31, 31–54. Lucas, J.W., Phelan, J.C., 2012. Stigma and status: the interrelation of two theoretical perspectives. Soc. Psychol. Quart. 75, 310–333. Marini, M.M., Fan, P., Finley, E., Beutel, A.M., 1996. Gender and job values. Sociol. Educ. 69, 49–65. Markowitz, F.E., 1998. The effects of stigma on the psychological well-being and life satisfaction of persons with mental illness. J. Health Soc. Behav. 39, 335–347. Markowitz, F.E., Angell, B., Greenberg, J.S., 2011. Stigma, reflected appraisals, and recovery outcomes in mental illness. Soc. Psychol. Quart. 74, 144–165. McGuffey, C.S., Rich, B.L., 1999. Playing in the gender transgression zone: race, class, and hegemonic masculinity in middle childhood. Gender Soc. 13, 608– 627. Messner, M.A., 2009. It’s all for the Kids: Gender, Families, and Youth Sports. University of California Press, Berkeley. Nunnally Jr., J.C., 1961. Popular Conceptions of Mental Health: Their Development and Change. Holt, Rinehart and Winston, New York. Olmsted, D.W., Durham, K., 1976. Stability of mental health attitudes: semantic differential study. J. Health Soc. Behav. 17, 35–44. Pascarella, E.T., Terenzini, P.T., 1991. How College Affects Students: Findings and Insights from Twenty Years of Research. Jossey-Bass, San Francisco. Pascoe, C.J., 2007. Dude, You’re a Fag: Masculinity and Sexuality in High School. University California Press, Berkeley.

A. Kroska et al. / Social Science Research 54 (2015) 68–79

79

Rashotte, L.S., Webster Jr., M., 2005. Gender status beliefs. Soc. Sci. Res. 34, 618–633. Ridgeway, C.L., 1997. Interaction and the conservation of gender inequality: considering employment. Am. Sociol. Rev. 62, 218–235. Ridgeway, C.L., Backor, K., Li, Y.E., Tinkler, J.E., Erickson, K.G., 2009. How easily does a social difference become a status distinction? Gender matters. Am. Sociol. Rev. 74, 44–62. Rosenfield, S., 1997. Labeling mental illness: the effects of stigma on life satisfaction. Am. Sociol. Rev. 62, 660–672. Rudman, L.A., 1998. Self-promotion as a risk factor for women: the costs and benefits of counterstereotypical impression management. J. Pers. Soc. Psychol. 74, 629–645. Rudman, L.A., Moss-Racusin, C.A., Phelan, J.E., Nauts, S., 2012. Status incongruity and backlash effects: defending the gender hierarchy motivates prejudice against female leaders. J. Exp. Soc. Psychol. 48, 165–179. Sanchez, L., Thomson, E., 1997. Becoming mothers and fathers: parenthood, gender, and the division of labor. Gender Soc. 11, 747–772. Sayer, L.C., 2005. Gender, time, and inequality: trends in women’s and men’s paid work, unpaid work, and free time. Soc. Forces 84, 285–303. Sayer, L.C., Bianchi, S.M., Robinson, J.P., 2004. Are parents investing less in children? Trends in mothers’ and fathers’ time with children. Am. J. Sociol. 110, 1– 43. Sears, D.O., 1986. College sophomores in the laboratory: influences of a narrow database on social psychology’s view of human nature. J. Pers. Soc. Psychol. 51, 515–530. Sears, D.O., 2008. College student-itis redux. Psychol. Inq. 19, 72–77. Sibicky, M., Dovidio, J.F., 1986. Stigma of psychological therapy: stereotypes, interpersonal reactions, and the self-fulfilling prophecy. J. Counseling Psychol. 33, 148–154. Stier, A., Hinshaw, S.P., 2007. Explicit and implicit stigma against individuals with mental illness. Aust. Psychol. 42, 106–117, http://dx.doi.org/10.1080/ 00050060701280599. Wright, E.R., Gronfein, W.P., Owens, T.J., 2000. Deinstitutionalization, social rejection, and the self-esteem of former mental patients. J. Health Soc. Behav. 41, 68–90.