Back to Journal Cover Page


CURRENT RESEARCH IN SOCIAL PSYCHOLOGY


Volume 7, Number 20
Submitted: April 26, 2002
First Revision: August 4, 2002
Second Revision: August 28, 2002
Accepted: August 28, 2002
Publication Date: August 28, 2002

IS GENDER STILL A STATUS CHARACTERISTIC?*

Rosemary Hopcroft
University of North Carolina at Charlotte

ABSTRACT

Past research has shown that gender acts as a diffuse status characteristic in social interaction. In particular, in a task situation males have been show to reject influence from females more than females reject influence from males. However, recent research suggests that the effects of gender may have changed, such that gender no longer acts as a diffuse status characteristic. In this experimental replication, results suggest that gender continues to act as a diffuse status characteristic in task situations.

[339]
---------------
[340]

Past research has established that gender works as a diffuse status characteristic in small group interactions, such that in an experimental task situation males reject influence from females more than females reject influence from males (Pugh and Wahrman 1983; Troyer 2001). Other studies also show this gender effect (e.g. Wagner Ford and Ford 1986; Foschi 1996; Foddy and Smithson 1999); although in all these studies the effect may appear stronger because subjects were given information that past research had shown that males performed better at the experimental task than females.

However, research results suggest that the effect of gender may have changed somewhat in recent years. For example, in one recent study of rejection of influence behavior in mixed-sex dyads, with no information provided about gender differences on task skill, Foschi and Lapointe (2002) found no effect of gender. Other studies have found evidence of recent change in the effects of gender. For example, Jasso and Webster (1997) found using 1974 data that there was a gender difference in earnings regarded as just for men and women, using 1995 data they found that this gender gap had not only closed but the earnings considered just for women were slightly higher than those considered just for men (Jasso and Webster 1999). Similarly, Foschi, Enns and Lapointe (2001) found that men were not more likely to reject influence than women in same-sex dyads, although a previous study had shown lower rejection of influence rates for the female dyads (Foschi and Freeman 1991). Rashotte and Smith-Lovin (1997) also found no difference in participation rates by males and females in mock jury groups as perceived by outside observers, although previous research had found a gender difference in participation in such groups (Smith-Lovin, Skvoretz and Hudson 1986).

Given these recent findings and evidence of possible change in the effects of gender, it becomes an important question whether, and where, gender continues to work as a diffuse status characteristic. An important issue is that experimental protocols differ across studies and can play an important role in producing effects. For example, as mentioned previously, those studies where participants are told that males have performed better at the experimental task in the past find stronger gender effects than studies where participants are not given this information (Wagner, Ford and Ford 1986; Foschi 1996). In addition, recently Troyer (2001; Troyer, Kalkhoff and Younts unpublished ms.) has found effects of both research protocol and the medium by which experimental information is communicated to the participants. Troyer (2001) found that explicitly telling participants that their partners disagreed with them as opposed to simply telling them what their partners chose made both the low and high status participants less inclined to reject influence, although it widened the gender gap. Additionally, Troyer et al. (unpub. ms.) showed that visual exposure to the "partner" by video increased the gender effect, probably by making gender salient for more participants.

Given these considerations, in the following study participants are not given any information to the effect that males have been shown to be better at the experimental task than females. Thus, the study should show the effect of gender as a diffuse status characteristic as it operates in the population. Second, participants were informed of their partners’ initial choices, but the disagreements were not stressed. That is, participants were not told explicitly whether their partners agreed or disagreed with them. Given Troyer's (2001) result, which was that explicit statement of agreement or disagreement had the effect of decreasing both participants’ tendency to reject influence, while at the same time increasing the gender gap in rejection of influence; then not including such an explicit statement may serve to mute the gender effect somewhat. Last, participants in this study did (very briefly) actually see their ostensible partners, as visual exposure to a partner is more likely to make gender salient than verbal or written information (Troyer et al. unpub. ms.). Further, in the mixed-sex dyads it eliminates any effects due to specific characteristics of the videotaped partner by randomizing the partner, although probably at the expense of somewhat higher variance. That is, in the mixed-sex dyads, using a single male or female videotaped partner opens up the possibility that any observed effects of the partner were due not to the partner's gender status, but to some other unspecified characteristic of that particular partner. Using multiple male and female partners, randomly assigned, removes the possibility of this problem. [1]

[340]
---------------
[341]

METHOD

Participants and Experimenters

Participants were 57 men and women aged 18-22 recruited from large classes at a moderately sized state university in the southeastern United States in the Summer and Fall of 2001. All participants were reimbursed $20 for their participation in the one-hour study. Participants were excluded from the study if they said they were Psychology majors or had experience of experiments involving deception. There were three conditions – 1) Men paired with women, 2) Men paired with men, 3) Women paired with men. That is, data were collected for each of three cells (out of a total of four possible) in the 2 (sex of subject) x 2 (sex of partner) factorial design. Data were collected for condition 2 in which gender was controlled to find the average tendency to reject influence in this population. Each experimental session was run by a female experimenter, and special attention was paid to ensure uniformity of experience for the participants across sessions. Post-task interviews and debriefings were performed by a male or female research assistant.

Procedures and Materials

Participants were scheduled in randomly assigned male/male or male/female pairs. Each participant worked on the experimental task alone in a small cubicle. Prior to being put in their cubicles, participants briefly saw their "partners." Participants were not given a chance to introduce themselves to their partners, or to talk or otherwise interact with their partners. The decision to use live partners was made in order to maximize the likelihood that gender would be salient. Moreover, because every participant saw a different partner, it eliminated the possibility that in the mixed-sex dyads some characteristic of a single videotaped partner (unrelated to gender) could influence the results in a systematic manner. That is, in the mixed-sex dyads, a single videotaped partner could have visible characteristics that were unforeseen by the researcher and unrelated to his or her gender but which unexpectedly promoted either rejection of influence or the opposite in participants. For example, the videotaped male may have a particular demeanor that promoted rejection of influence by female partners, or the videotaped female may have a demeanor that promoted the opposite by male partners. Randomizing partners avoids this problem. After seeing their partners, participants were escorted to separate small rooms. Each room was equipped with a video screen, computer monitor, mouse, and closed circuit camera. The equipment was pointed out to him or her, then each participant was asked to fill in a sign-in questionnaire on the computer monitor. The purpose of this sign-in questionnaire was to acquaint the participants with the use of the mouse and computer. 

Next, participants were shown a video featuring Dr. Philip Gordon, who told them that they were participating in a study of contrast sensitivity.[2] Participants were told that some individuals have higher levels of contrast sensitivity than others, although contrast sensitivity is not related to any more familiar ability such as artistic ability or math ability. The video then informed participants about the contrast sensitivity task. This task consisted of choosing which of a pair of slides containing black and white rectangles contained the greater amount of white area. Each participant was told that he or she would make a decision, and then would be shown his or her partner’s decision, and then, after taking into consideration his or her partner’s choice, he or she would make a final decision. Each participant was then presented with a practice slide for the purpose of acquainting them with the procedure for choosing a slide, seeing his or her ostensible partner's choice, and making a final decision. Participants were also introduced to the concept of team scoring by Dr. Gordon's assistant, Ms. Lynn Mason. That is, they were told for each correct individual final answer, the team would receive one point. The participants were told that their goal was to maximize their team's point score. The team scoring concept is often used in such studies (e.g. Foschi and Lapointe 2002) as it helps ensure that participants are task-focused (see below). However, unlike in Foschi and Lapointe’s (2002) study, here participants were also introduced to the concept of team standards, that is, the distribution of scores ostensibly found in the past. [3] That is, they were told that 35-40 is a superior team score, 31-34 is an above average team score, 22-30 is an average team score, 18-21 is a below average team score, and 0-17 is a poor score. Some version of this team standards manipulation is often used in studies of this sort to increase task focus in participants (e.g. Moore 1968; Troyer 2001).

[341]
---------------
[342]

After the instructions, participants were presented with a series of 23 slides on the computer. For each slide, participants made an initial choice, and then were informed, also via the computer, of their partners' ostensible choices. The computer was programmed so that on 20 out of the 23 slides the partner apparently disagreed with the participant. The participant then made a final choice. Rejection of influence, or P(s), probability of staying, was measured as the proportion of the 20 disagreements that the participant stuck with her or his initial choice and did not change to the partner's ostensible initial choice.

After they finished the task participants were asked to complete another questionnaire in order to determine, first, if the two key scope conditions for the experiment were met. That is, the participants needed to have been both task-focused and collectively-oriented (Foschi 1997). That is, they had to be trying to get the correct answer and they must have considered the opinions of their partners before making their final decisions. Second, the questionnaire verified that each participant was aware of his or her partner's gender. Last, the questionnaire also served to help find out if a participant was suspicious to the extent that it changed her or his behavior on the task. All these issues were probed more deeply in subsequent individual interviews, after which participants were debriefed.

RESULTS

On the basis of information from the final questionnaire and the exit interviews, 1 woman out of a total of 19 were excluded, and 6 men out of a total of 45 were excluded. That is, overall 7 out of 64, or 11 percent of the participants, were excluded from the analysis. Rules for exclusion were formulated ahead of time and were based primarily on whether the scope conditions of task focus and collective orientation were met. The most common reasons for exclusion were: a high level of suspicion of the experiment that the respondent said influenced how he or she performed the task (e.g. not taking it seriously), ignoring the partner's choice, and misunderstanding or ignoring of directions. Nevertheless, the overall exclusion rate is low compared with the exclusion rates from other similar experiments (e.g. Berger et al. 1992; Troyer and Younts 1997; Foschi et al. 2001).

Results were analyzed using a t-test of independent samples. The Levene test for equality of variance was significant, so a t-test assuming unequal variances was used. For the women paired with men, average P(s) was 0.53, for men paired with women, average p(s) was 0.64. This difference was significant, p equals 0.007. That is, the men rejected influence from the women more than the women rejected influence from the men. Ordinal predictions from status characteristics theory (Berger et al. 1992) were also confirmed, that is, P(s) condition 1 greater than P(s) condition 2 greater than P(s) condition 3. These P(s) values are slightly lower (but not significantly lower) than those predicted using Fisek et al. (1992) for this situation, that is, where actors gain or lose advantage from a single diffuse status characteristic, in this case gender, and very close to those found by Troyer (2001) in a similar study. Troyer found in her baseline study P(s) equals 0.55 for women paired with men, and P(s) equals 0.65 for men paired with women.

[342]
---------------
[343]

Table 1: Rejection of Influence by Condition

Condition
Subject (partner)

 

N

Predicted P(s)
(Fisek et al.1992)

Average P(s)
(0=no rejection,
1=full rejection)

SD

 

1. Male (female)

 

15

 

.65

 

.64

 

. 09

 

2. Male (male)

 

24

 

.60

 

.61

 

.17

 

3. Female (male)

 

18

 

.55

 

.53

 

.14

DISCUSSION AND CONCLUSION

Paraphrasing Mark Twain, reports of gender's demise as a status characteristic may have been exaggerated. The results of this study are in line with those from Troyer (2001) and Troyer et al. (unpub.ms) and previous studies and suggest that gender lives on as a diffuse status characteristic. That is, in an experimental task situation where no information linking the task to gender was given, and where participants briefly saw their ostensible partners, males rejected influence from females more than females rejected influence from males.

These results were found with a group of young 18-22 year olds at a medium-sized public university in the southeast. It may be that there is something about this population (e.g. southern culture) that explains why gender persists as a status characteristic in this population and not in a similar population in the Pacific Northwest (Foschi and Lapointe 2002). Regional variations in the effect of gender as a status characteristic are a real, under-explored possibility. Further, as Troyer (2001) and Troyer et al. (unpub.ms) have shown, experimental protocol does influence results and there may be some aspects of the experimental protocol here that made the gender effect more likely. For example, it is likely that visual exposure to an opposite sex partner (the protocol used here) made gender salient to more participants (see Troyer et al. unpub. ms.) than the protocol of simply being verbally told the gender of the partner, as was the case in the Foschi and Lapointe (2002) study. On the other hand, Troyer (2001) found a similar gender effect to that reported here with the status of the partner (male or female) given to participants in writing only, and it is hard to see why information in writing would make gender salient more often than verbal information, although this is an as yet unexplored possibility.

[343]
---------------
[344]

Alternatively, it may be the lack of the "team standards" manipulation in the Foschi and Lapointe (2002) protocol that accounts for the discrepancy in findings. The current study and both the Troyer studies use this manipulation. Troyer (2001) found that a brief discussion of performance scores (compared to a detailed discussion of them) did reduce levels of task focus in participants, so it may be that no discussion of performance scores at all reduced levels of task focus further in participants in Foschi and Lapointe (2002), and hence also reduced the gender effect as measured by P(s). However, this seems unlikely, given that Foschi and Lapointe were careful to ensure that all included participants were task-focused. Alternatively, given that Foschi and Lapointe (2002) found depressed P(s) for males only (P(s) equals 0.54); it may be that males, and not females, were influenced by the lack of the "team standards" manipulation. Resolution of these unexplored issues is beyond the scope of the present study and is a task for future research.

Whatever the reason for the null finding in Foschi and Lapointe (2002), given that the majority of previous studies, including very recent studies such as Troyer (2001), Troyer et al. (unpub. ms.), in addition to the current study, found the gender effect suggests that the most conservative conclusion is that gender continues to operate as a status characteristic in task situations.

ENDNOTES

[1] Assuming a particular videotaped partner has identical effects on both male and female participants, one can control for videotaped partner effects by comparing male and female average p(s) when paired with the same male videotape partner, and by comparing male and female average p(s) when paired with the same female videotape partner. This gives us male/male compared to female/male comparison, and the female/female compared to male/female comparison. The problem remains however, if we attempt to compare female/male average p(s) with male/female average p(s), as is the case here. In this case, any difference in average p(s) between female/male and male/female pairs may be due to some unknown quality of either or both of the videotaped opposite sex partners.

[2] This videotape is a somewhat abbreviated version of a tape created by Professor Murray Webster et al. in 2001 for another expectation states research project. The videotape was edited to contain only the sections necessary for the experiment reported here. Many thanks are due to Professor Webster for his provision of this tape.

[3] In previous studies, Foschi (e.g. 1996) has not used the concept of team scoring because her purposes were to examine the emergence of double standards. Presenting one set of standard team scores may have undermined this effort.

[344]
---------------
[345]

REFERENCES

Berger, Joseph, Robert Z. Norman, James W. Balkwell, and Roy F. Smith. 1992. "Status Inconsistency in Task Situations: A Test of Four Status Processing Principles." American Sociological Review 57: 843-855.

Fisek, M. Hamit, Robert Z. Norman, and Max Nelson-Kilger. 1992. "Status Characteristics and Expectation States Theory: A Priori Model Parameters and Test." Journal of Mathematical Sociology 16:285-303.

Foddy, Margaret and Michael Smithson. 1999. "Can Gender Inequalities be Eliminated?" Social Psychology Quarterly 62, 4: 307-324.

Foschi, Martha. 1996. "Double Standards in the Evaluation of Men and Women." Social Psychology Quarterly 59,3: 237-254.

Foschi, Martha. 1997. "On Scope Conditions." Small Group Research 28, 4: 535-555.

Foschi, Martha, Sandra Enns and Vanessa Lapointe. 2001."Processing Performance : Evaluations in Homogeneous Task Groups: Feedback and Gender Effects." Advances in Group Processes 18: 185-216.

Foschi, Martha and S. K. Freeman. 1991. "Inferior performance, standards and influence in same-sex dyads." Canadian Journal of Behavioural Science23: 99-113.

Foschi, Martha and Vanessa Lapointe. 2002. "On Conditional Hypotheses and Gender as a Status Characteristic." Social Psychology Quarterly 65, 2: 146-162.

Jasso, Guillermina; Webster, Murray, Jr. 1997. "Double Standards in Just Earnings for Male and Female Workers." Social Psychology Quarterly, 60, 1: 66-78.

[345]
---------------
[346]

Jasso, Guillermina; Webster, Murray, Jr. 1999. "Assessing the Gender Gap in Just Earnings and Its Underlying Mechanisms." Social Psychology Quarterly 62, 4: 367-380.

Moore, James C. Jr. 1968. "Status and Influence in Small Group Interactions."Sociometry. 31: 47-63.

Pugh, M. D. and Ralph Wahrman. 1983. "Neutralizing Sexism in Mixed-Sex Groups: Do Women Have to Be Better than Men?" American Journal of Sociology 88, 4: 746-62.

Rashotte, Lisa Slattery and Lynn Smith-Lovin (1997). "Who Benefits from Being Bold: The Interactive Effects of Task Cues and Status Characteristics on Influence in Mock Jury Groups." Advances in Group Processes. 14:235-255.

Smith-Lovin, Lynn, John V. Skvoretz and Charlotte G. Hudson. 1986. "Status and Participation in Six-Person Groups: A Test of Skvoretz's Comparative Status Model." Social Forces 64:992-1005.

<>Troyer, Lisa. 2001. "Effects of Protocol Differences on the Study of Status and Social Influence." Current Research in Social Psychology. http://www.uiowa.edu/~grpproc/crisp/crisp.6.13.htm

Troyer, Lisa, Will Kalkhoff and C. Wesley Younts. 2002. "Effects of Media and Gender on Status-Organizing Processes." Unpublished manuscript, University of Iowa.

Troyer, Lisa and C. Wesley Younts. 1997. "Whose Expectations Matter? The Relative Power of First and Second Order Expectations in Determining Social Influence." American Journal of Sociology. 2: 692-732.

Wagner, David G., Rebecca S. Ford and Thomas W. Ford. 1986. "Can Gender Inequalities be Reduced?"American Sociological Review 51: 47-61.

ACKNOWLEDGEMENTS

The author would like to thank Murray Webster for the use of the Group Processes Lab and equipment, including videos and software for this study; and for his assistance with experimental design and recruiting. More than is typically the case, this study could not have been done without him. Of course, any problems are my own doing. I would also like to thank the anonymous reviewers and Martha Foschi for helpful comments, Joe Whitmeyer for comments and research assistance, and Rebecca Jing Tang, and Stephanie Southworth-Brown for research assistance. This work was funded in part by a University of North Carolina at Charlotte Faculty Research Grant.

AUTHOR BIOGRAPHY

Rosemary L. Hopcroft is an Associate Professor of Sociology at the University of North Carolina at Charlotte. She is interested in explaining gender inequalities using sociological (and other) theories, and is currently involved in an experimental study examining the interaction between age and gender on influence. She has also published extensively in the area of comparative and historical sociology. E-mail: rlhopcro@email.uncc.edu

[346]
---------------
[347]