Back to Journal Cover Page


CURRENT RESEARCH IN SOCIAL PSYCHOLOGY


Volume 5, Number 11
Submitted: October 21, 1999
Resubmitted: April 4, 2000
Accepted: April 18, 2000
Publication date: April 24, 2000

STATUS CHARACTERISTICS THEORY AND SEXUAL ORIENTATION: EXPLAINING GENDER DIFFERENCES IN RESPONSES TO SEXUAL ORIENTATION

Kevin Childers
Louisiana State University

ABSTRACT

In this paper I consider ways in which sexual orientation, either homosexual or heterosexual, affects group structure and individual responses to others. In particular I test whether men, compared with women form lower performance expectations for homosexual targets. I then test two mechanisms that may produce that result — universal salience and weighting. This research results from the status characteristics theoretical tradition explicated in Berger and Zelditch (1993). This study uses data from two vignette studies, Webster and Hysom (1995) and Webster, Hysom, and Fullmer (1998). Results show that men respond more negatively to a homosexual sexual orientation than women. Further, results do not show clear support for either proposed mechanism to explain these differences.

[191]
---------------
[192]

INTRODUCTION

Research in status characteristics theory supports the assertion that sexual orientation, homosexual and heterosexual, is a diffuse status characteristic in The United States (Johnson 1995; Webster and Hysom 1995; Webster et al. 1998). Separately, stereotyping research indicates that men respond more negatively to a homosexual sexual orientation than women (Garnets and Kimmel 1993; Herek and Glunt 1993; Martin 1990; Whitley and Kite 1995). There is also empirical support within the status characteristics tradition for men and women not always responding similarly to status characteristics (Ridgeway, Johnson, and Diekema 1994; Smith-Lovin and Brody 1991).

Based on these two research programs, status characteristics theory and stereotyping theories, I explore more precise mechanisms of the expectation formation process for sexual orientation. I do two things here. First, I test whether men form lower expectations for homosexuals, compared with heterosexuals, than do women. Second, I try to determine what mechanism causes such expectation differences. I suggest, and test, two theoretical mechanisms for why men might form lower expectations than women. The first mechanism, universal salience, assumes that sexual orientation is not a salient characteristic for women as it is for men. The second mechanism, weighting, assumes that sexual orientation is salient for both men and women, but that men weigh sexual orientation more heavily than women in forming expectations.

STATUS CHARACTERISTICS THEORY

Status characteristics theory describes a social psychological process by which people come to hold performance expectations for themselves and others (Berger and Zelditch 1993). When people interact, they bring with them personal characteristics called status characteristics. Examples of status characteristics in The United States are education level, race, gender, age, and physical attractiveness. People use status characteristics in evaluating their work and that of others and they shape interactions in many ways that vary depending on features of an interaction.

[192]
---------------
[193]

In groups that are task and collectively oriented, status characteristics that differentiate actors or that actors believe relevant to the task may become important in determining task expectations for self and other. Unless evidence exists to the contrary, actors treat salient status characteristics as if they are relevant to task performance, even if no reasonable connection between ability and the status characteristic(s) exist. Actors use all salient status information in evaluating individual performance, and combine information either through attenuation or consistency (Berger and Zelditch 1993). Expectations become observable behavior in groups such as amount of talk, opinion giving and accepting, suggestion giving or accepting, and choice of group leader.

To explore the application of status characteristics theory to sexual orientation, research must show that sexual orientation, homosexual and heterosexual, meets the theoretical definition of a status characteristic. That is, research must show that 1) sexual orientation has at least two, differentially evaluated states, homosexual and heterosexual, and that 2) each state has sets of specific and general performance expectations.

Johnson (1995) reviews status characteristics, stereotyping, and attitudinal literature about homosexuals. Johnson found strong support for the idea that there are at least two states of sexual orientation which people evaluate differentially, heterosexual and homosexual. She also found strong support that heterosexual is the more positively evaluated state in the United States. Research by others (Herek 1991; Preston and Stanley 1987), provide additional evidence that sexual orientation has at least two differently evaluated states.

Studies by Webster and colleagues (Webster and Hysom 1995; Webster, Hysom, and Fullmer 1998) provide a test of point number two above: that there exist specific and general performance expectations associated with each state of sexual orientation. Both studies intended to establish whether sexual orientation is a diffuse status characteristic in the United States.

The researchers use pairs of vignettes with one homosexual and one heterosexual target. One study contained male targets while the other contained female targets. Researchers followed vignettes with a series of questions to assess, among other things, expectations for each target at certain tasks, such as leadership ability or abilities at tasks in general.

[193]
---------------
[194]

Data from both Webster and colleagues’ studies support that there are differential expectations associated with each state of sexual orientation. Further these differential expectations were in the predicted direction. That is respondents, in general, assigned higher performance expectations to the heterosexual target compared with the homosexual target.

HYPOTHESES

I have two research questions. First, do male respondents form lower expectations for homosexual targets than for heterosexual targets, relative to female respondents in the United States. The addition of this question to existing work is an examination of gender differences in expectation formation based on sexual orientation. Stereotyping research suggests that men and women do respond differently to a heterosexual sexual orientation compared with a homosexual sexual orientation (Garnets and Kimmel 1993; Herek and Glunt 1993; Martin 1990; Whitley and Kite 1995). My hypothesis is that men will form lower expectations for both gay male and lesbian targets on the general expectations scale than will women.

My second hypothesis a possible mechanism for a difference in women’s and men’s responses. This hypothesis examines variance about the mean for men’s and women’s responses on the extended general expectation states questions. For this reason, it is necessary at this point to explain a little bit about the questions and the scale upon which the questions are scored.

The expectation states scale is a measure incorporating responses to several questions (for a list of these questions see Webster and Hysom 1995). All questions are on a seven point scale. The scale is scored from 1.0 to 7.0 with equality being 4.0. Questions, when combined, reflect respondent’s expectations for targets’ abilities at tasks in general.

The argument, presented above, that 4.0 represents equality on this scale is consistent with work by others using this or similar scales. The assumption is that if people do not form differential expectations based on the characteristic of interest, then the average of their responses to questions assessing predicted task ability for two targets differing only on the characteristic of interest will be approximately at the center of the scale.

[194]
---------------
[195]

This is also in keeping with the general argument of status characteristics theory that only those characteristics people believe important or relevant to a given task become salient and therefore affect expectations. In the vignette study, the only basis upon which to differentiate targets is based on sexual orientation. So, if respondents do not activate sexual orientation then their mean responses for the targets should be 4.0, or neutral on the scale.

Based on my first hypothesis, on the scale men’s responses would displace further from equality than women’s. If my hypothesis is supported a further question, and my second research question, is why this occurs? That is, what theoretical mechanism could explain why men form lower expectations for gay and lesbian targets than women?

I envision two mechanisms that could produce the difference between men and women. The first mechanism is universal salience. The second mechanism is weighting. Each carries different interpretations for the social meaning of sexual orientation.

The universal salience mechanism depends on the first assumption of the status generalization process: salience. This assumption states that if a status characteristic differentiates actors, or is perceived relevant to the group task, that status characteristic will become important in forming performance expectations for self and other. This mechanism, then addresses the issue of whether a characteristic becomes salient equally for everyone. That is, it is possible that for certain subgroups, sexual orientation may not be a salient status characteristic. Some researchers have found that for certain subgroups of college age women, gender may no longer be a salient status characteristic (Ridgeway et al. 1994; Smith-Lovin and Brody 1991; Stewart 1988). In the case of these subgroups, when gender differentiates actors, they will not perceive gender as salient and they will not, therefore, use gender as a factor upon which to base performance expectations. If this does not occur then the status generalization process does not occur in these groups.

[195]
---------------
[196]

It is plausible that for some subgroups, sexual orientation is not a salient status characteristic. In the case of these people, if presented with a situation where only sexual orientation differentiates actors, as with the vignettes in the Webster and colleague studies, then the status generalization process would not occur. The general expectation states scale mean responses of such a subgroup would not differ significantly from 4.0. Specifically here, if men show a more negative response to a homosexual sexual orientation than women, then women are more likely to be a subgroup for whom sexual orientation is not salient. It is then possible to test this idea based on the mean response of women and whether that general expectation states scale mean response is significantly different from 4.0.

This would actually be testable by the variance around the means of these groups. Men, as a group, would have scores that are tightly distributed around the mean. Women, however, would have a widely distributed variance. This wide distribution would suggest that women do not respond uniformly, as a group, to sexual orientation. Further, this suggests that women fall into two or more groups. One possible classification would be a group for whom sexual orientation does not matter in forming expectations for task behavior and another group for whom sexual orientation does matter.

The other mechanism, weighting, assumes that sexual orientation becomes salient and that the status generalization process occurs. However, this mechanism assumes that men and women weigh sexual orientation differently. In this case, all respondents notice sexual orientation and notice that it differentiates actors and is therefore used as a salient dimension upon which they may base performance expectations. The key is that one group views sexual orientation as more important in determining performance than another. In terms of this study, men are likely to weight sexual orientation more strongly in forming performance expectations than are women.

The weighting mechanism addresses a simplification in status generalization theories. Research has not examined this mechanism before because the theories as formulated have been successful at predicting observable behavior. Status generalization theories treat all diffuse status characteristics as if they were equally important. Status generalization theories have focused on the degree of relevance to task performance of a given characteristic not its weight.

[196]
---------------
[197]

Intuitively, believing that some characteristics are more important than others in forming performance expectations is plausible. In other words, according to the theory, once the status generalization process has been activated for a given group, each actor for whom sexual orientation is salient will react similarly. So the second mechanism suggests that the differences between women’s and men’s responses is in the relative importance of the characteristic.

To repeat and expand on what I said above, in terms of data, I can examine the first question, whether men and women differ in their responses to sexual orientation, in a straightforward manner. Comparing the mean general expectation states scale scores of men to those of women and determining if those means differ significantly answers this question. This should, however, be done separately for each Webster and colleagues’ study since the gender of vignette targets vary for the two Webster and colleagues’ studies. If gender differences in responses to a homosexual sexual orientation occur then additional analyses could determine which mechanism is responsible.

If the mechanism is salience, the vignettes produce two subgroups of women. One subgroup exists for whom sexual orientation is salient and therefore forms differential expectations. Another subgroup exists that does not treat it as salient and therefore has no reason to form differentiated expectations. The mean general expectation states scale score of female respondents then, is a combination of the two subgroups. In contrast, under this mechanism, we assume sexual orientation is salient for nearly all male respondents.

If this situation produced two subgroups among female respondents, evidence for it might be found by examining the variance of women’s responses. The presence of subgroups in a condition will increase the variance of respondents, so variance is often used as a test for bimodality. Support for the salience hypothesis would appear if variance for women were considerably greater than for men. (Because variances are distributed normally, if variance in one condition is 1.96 times greater than variance in a second condition the conditions differ significantly.) Support for the weighting hypothesis would require approximately equal variances for women and men, or even greater variance for men.

[197]
---------------
[198]

DATA AND RESULTS

In the Webster and colleagues’ studies (1995; 1998) there were four conditions. For several reasons I analyze only conditions one and two in this paper. Condition one manipulated only sexual orientation. That is, in the vignettes, targets varied only by their sexual orientation. In the first study, one target was a gay male, while the other target was a heterosexual male. In the second study, one target was a lesbian, while the other was a heterosexual female.

Condition two manipulated occupation. That is, in the vignettes, targets differed only by their occupation. One target was a dishwasher (low status occupation), the other target was a computer systems analyst (high status occupation). Webster and colleagues intended this condition to test the measure itself by providing data on expectation formation based on occupation that is comparable to previous findings for occupation.

In the Webster and Hysom study (1995), there were 269 usable questionnaires (Nfemales = 157; Nmales = 108; Nnonresponding = 4). Condition 1 had 71 respondents, condition 2 had 73, condition 3 had 61 respondents, and condition 4 had 64 respondents.

In Webster et al. (1998) there were 549 usable questionnaires (Nfemales = 351; Nmales = 188; Nnonresponding = 10). Condition 1 had 150 respondents, condition 2 had 132 respondents, condition 3 had one 136 respondents, and condition 4 had 131 respondents.

Given that the two Webster and colleagues’ studies support sexual orientation as a diffuse status characteristic, I can now examine gender differences. If gender differences appear then analyses to determine the mechanism producing those differences are possible.

[198]
---------------
[199]

Table 1. Study 1, Means, Standard Errors, and Variances Across Conditions By Gender

Condition
(N)

Extended General Expectation States & (S.E.) For Males

Variance For Males

Extended General Expectation States & (S.E.) For Females

Variance For Females

1
(M = 27; F = 43)

3.6281 (.1296) .4534 3.9643 (.0897) .3381

2
(M = 32; F = 40)

3.0217 (.1716) .8830 3.0944 (.1804) 1.2041

 

Table 2. Study 2, Means, Standard Errors, and Variances Across Conditions By Gender

Condition
(N)

Extended General Expectation States & (S.E.) For Males

Variance For Males

Extended General Expectation States & (S.E.) For Females

Variance For Females

1
(M = 48; F = 100)

3.9004 (.0781) .2808 3.9184 (.0503) .2404

2
(M = 54; F = 76)

2.8726 (.1049) .5721 3.0211 (.0900) .5753

In study 1, both conditions 1 and 2 male respondents form lower expectations for the low status target than female respondents. This means that the response mean for men has a greater negative displacement from 4.0 (the scale’s center) than the female response mean. Male respondents produced a mean general expectation states scale score of 3.63 with a standard error of .1296. This score represents a displacement from 4.0 of -.3719. Female respondents produced a mean general expectation states scale score of 3.96 with a standard error of .0897 which represents a displacement of -.0357. A t-test showed the difference in means to be significant at the .05 confidence level, and thus results for study 1 in condition 1 support the hypothesis that men form lower expectations for homosexual targets. There is no significant difference between men and women in variance. These results are consistent with the weighting hypothesis.

[199]
---------------
[200]

There is no significant difference in variances because, as I discussed earlier, to be significant one variance must be at least 1.96 times as large as the other. With a variance of .4534 for men and .3381 for women it is clear one variance is not 1.96 times as large as the other. These results are not consistent with the salience hypothesis because the variances are not significantly different. In fact, the variance for men is larger than that for women which is the opposite of what was predicted.

Study 2 produced no significant difference in men’s and women’s responses in condition 1. The mean response for men is 3.9004 while the mean response for women is 3.9184. Since there is no significant difference between male and female respondents there is no reason to examine variance. Results for study 2 then are not consistent with either the salience or the weighting hypothesis. For study 1 the hypothesis that men would form lower expectations for homosexual targets than women is supported while in study two it is not.

Study 1 produced results showing that, for men, sexual orientation is clearly a diffuse status characteristic, carrying positive and negative evaluations, and also specific and general performance expectations. For women, however, sexual orientation may or may not be a status characteristic. The results do not show significant displacement from 4.0. It is possible that this is due to the lower number of respondents than when men and women are examined together.

DISCUSSION

In this paper I use data from Webster and Hysom (1995) and Webster, Hysom, and Fullmer (1998) to explore two research questions. First, do men form more negative beliefs about homosexuals than do women. In terms of the data, men would displace further from the mean response on the study indicators than would women. Second, if this difference in men’s and women’s responses exists, what mechanism produces the difference. I propose two theoretical mechanisms, salience and weighting, which may explain the differences.

I find empirical support for the idea that men form more negative beliefs about homosexuals than do women. This finding provides theoretically sound, empirical evidence to support findings by stereotyping researchers (Garnets and Kimmel 1993; Herek and Glunt 1993; Martin 1990; Whitley and Kite 1995).

[200]
---------------
[201]

Research by Webster and Hysom (1995) and Webster et al. (1998) show that sexual orientation functions as a diffuse status characteristic in United States culture. The performance expectations people form for gays and lesbians, as for other status characteristics, create both status advantages (for heterosexuals) and disadvantages (for lesbians and gay men) in United States culture. Further, those status advantages and disadvantages structure interactions in small collective and task oriented groups and predictably affect observable behaviors.

Webster and Hysom (1995) and Webster et al. (1998) explain a previously unexplained aspect of the social meaning of sexual orientation. That aspect is that sexual orientation produces performance expectations and carries status advantages and disadvantages for both women and men. It should be understood that status effects are not the only factors shaping the social meaning of sexual orientation.

Status-based aspects of the social meaning of sexual orientation explain why lesbian or gay people may not get raises or promotions among other things. It also helps us to know some ways to overcome some undesirable effects of status generalization. One possible application, dealt with in this study, is the issue of a gay person’s leadership of civilian and combat units in the armed forces. Under certain conditions, how others perceive ability to command when sexual orientation is known or inferred, can be predicted. A person’s sexual orientation may also affect the amount of respect a gay commander receives from subordinates. We can adapt intervention strategies from previous status-based research on overcoming status effects (E. G. Cohen 1993; Cohen and Catanzarite 1988; and Entwisle and Webster 1974).

The second issue addressed in this paper is an oversimplification in status generalization theories. The issue of weighting is as intriguing as the issue of salience. Weighting is the idea that a characteristic may be salient for nearly everyone, but some people consider it more heavily in determining performance expectations. Status generalization theories have thus far ignored the idea that characteristics may weight differently. If it can be determined that they do, then it is a logical extension of the theories to determine which characteristics weight more heavily and the extent to which one characteristic weights more than another.

[201]
---------------
[202]

Results here do not show strong support for either the salience or the weighting hypothesis. In study one, salience is supported, but in study two, neither is supported. Since the respondents are similar for each study and are randomly selected, as are the questionnaires similar except for the use of lesbian, instead of gay male, targets in the vignettes, I am led to the conclusion that neither mechanism is supported.

In both studies the results could be stronger. It is possible that self presentation concerns presented problems. Some respondents may hesitate to express their true responses on the questionnaires for fear they may appear homophobic. With intense multicultural efforts on many college campuses, and in some high schools, it is possible that differential evaluations of any kind have been discouraged.

It is also possible that the lack of difference in study 2 is a result of gender of target. The purpose of doing the second study was to test whether status generalization based on sexual orientation was also linked to gender of target. The fact that we see stronger effects for sexual orientation in study 1 than in study 2 might show the respondents are reacting to issues of both gender conformity and sexual orientation.

Cathryn Johnson (1995) discusses this idea in her literature review and codification. She proposed that knowledge of a person’s homosexual orientation may trigger salience of a second status characteristic: gender. She notes that people often associate gender non-conformity with homosexuality. For instance, many people assume gay men are also feminine, thus activating gender while simultaneously activating sexual orientation. If confirmed, this would explain cases of unusually strong interaction disadvantages sometimes present when both homosexuality and gender non-conformity are observed. It could also guide investigations of the components of this status characteristic.

[202]
---------------
[203]

Two areas for further research arise. The first is to examine interaction effects of sexual orientation of respondent with sexual orientation of targets. In other words, researchers could examine whether lesbian and gay respondents form expectations the same as heterosexual respondents. Since the status process is not voluntary nor conscious, we would assume that performance expectations would form among both subgroups . This would seem to be as true for sexual orientation as it is for other status characteristics. However, recent research by Ridgeway et al. (1994) may provide a basis for the opposite argument. In her study, Ridgeway found that gender no longer is a status characteristic for at least some college age women. Also, Smith-Lovin and Brody (1991) found that men interrupt women in conversation more often than they interrupt other men, while women interrupt women and men equally. It is possible, from these two examples, that sexual orientation may not function equally for heterosexuals and lesbians and gays.

Second, it would seem logical from this point to apply research on sexual orientation as status to actual interaction sets. Webster and Hysom (1995) and Webster et al. (1998) have established, preliminarily, that sexual orientation of male and female targets affects performance expectations. There may be added effects when a multidimensional person is represented and not just a questionnaire target. We may find it useful to add to a person’s sexual orientation physical mannerisms, style of dress, and other personal characteristics. This type of study would most likely clear up questions about the findings of the first two studies. We could use interaction groups to address Johnson’s (1995) question of the activation of gender nonconformity and sexual orientation.

In summary, research now exists that gives evidence that sexual orientation -- homosexual or heterosexual -- is a diffuse status characteristic in United States culture. It has further been shown that in certain cases men and women form different performance expectations based on sexual orientation. Those performance expectation differences are due to the view by various respondents of sexual orientation as salient. This adds to the understanding of the status generalization process. It also adds to the understanding of intervention strategies to overcome discrimination and prejudice based on sexual orientation. Other research could further expand knowledge of this aspect of the social meaning of sexual orientation or expand its theoretical applications.

[203]
---------------
[204]

REFERENCES

Berger, Joseph and Morris Zelditch. (1993). Theoretical Research Programs: Study in the Growth of Theory. Stanford, CA: Stanford University Press.

Cohen, Elizabeth G. (1993). "From theory to practice: The development of an applied research program." Pp. 385 415 in J. Berger and M. Zelditch, Jr. (Eds.), Theoretical Research Programs: Studies in the Growth of Theory. Stanford, CA: Stanford University Press.

Cohen, Elizabeth G., and Lisa Catanzarite (1988). "Can expectations for competence be altered in the classroom?" Pp. 27 - 54 in Murray Webster Jr., and Martha Foschi (eds.), Status Generalization: New Theory and Research. Stanford, CA: Stanford University Press.

Entwisle, Doris and Murray Webster, Jr. (1974). "Raising children’s expectations for their own performance: A classroom application." Pp. 211 - 243 in Joseph Berger, Thomas L. Conner, and M. Hamit Fisek (eds.), Expectation States Theory: A Theoretical Research Program. Cambridge, MA: Winthrop.

Garnets, Linda D., and D. C. Kimmel (1993). "Lesbian and Gay Male Dimensions in the Psychological study of Human Diversity." Pp. 1 - 52 in Linda Garnets and D. C. Kimmel (eds.), Psychological Perspectives on Lesbian and Gay Male Experiences.

Herek, Gregory M. (1991). "Stigma, prejudice, and violence against lesbians and gay men: Correlates and gender differences." Pp. 60 - 80 in J. Gonsiorek and J. Weinrich (eds.), Homosexuality: Research Implications for Public Policy. Newbury Park, CA: Sage.

Herek, Gregory M., and Eric K. Glunt (1993). "Interpersonal contact and heterosexuals’ attitudes toward gay men: Results from a national survey," The Journal of Sex Research, 30:239-244.

[204]
---------------
[205]

Johnson, Cathryn (1995). "Sexual orientation as a diffuse status characteristic: Implications for small group interaction." Pp. 115 - 137 in B. Markovsky, M. Lovaglia, and K. Heimer (eds.), Advances in Group Processes, Vol. 12. Greenwich, CT: JAI Press.

Martin, Carol L. (1990). "Attitudes and Expectations about Children with Nontraditional and Traditional Gender Roles." Sex Roles, 22:151-165.

Ridgeway, Cecilia L., Cathryn Johnson, and David Diekema (1994). "External status, legitimacy, and compliance in male and female groups." Social Forces, 72:1,051-1,077.

Smith-Lovin, Lynn, and C. Brody (1991). "Interruptions in group discussions: The effects of gender and group composition," American Sociological Review, 54: 424-435.

Stewart, Penni (1988). "Women and men in groups: A status characteristics approach to interaction." Chap. 4, pp. 69 - 85 in Webster, Murray, Jr., and Martha Foschi (eds.), Status Generalization: New Theory and Research. Stanford, CA: Stanford University Press.

Webster, Murray, Jr., and James E. Driskell (1983). "Beauty as Status," American Journal of Sociology, 89: 140-165.

Webster, Murray, Jr., Stuart Hysom, and Elise Fullmer (1998). "Sexual Orientation and Occupation as Status." Advances in Group Processes, by E. J. Lawler (ed.).

Webster, Murray Jr., and Stuart Hysom (1995). "Orientation as Status." Paper presented at the annual meetings of the American Sociological Association. Washington, DC.

Whitley, Bernard E., Jr., and Mary E. Kite (1995). "Sex differences in attitudes toward homosexuality: A comment on Oliver and Hyde (1993)." Psychological Bulletin, 117:146-154.

[205]
---------------
[206]

AUTHOR BIOGRAPHY

Kevin Childers is a doctoral candidate at Louisiana State University working under the direction of Professor Dawn Robinson. His dissertation is titled, "Exploring Humor as a Status Behavior in Task Groups: Making the Case for a Sociological Theory of Humor for Small Groups Research." His research interests include status characteristics theory, affect control theory and the sociology of emotions. His email address is unccgrad@aol.com.

[206]
---------------

Back to Journal Cover Page