.I
‘£

$i
-
-4

*9

”394"“

~ .
~ “V‘sfifi‘

mum

I

W

k- a

L¢.:zu:

 

 

 

anythd'g‘h u ~ u-rgmneuvvm mtfifimetfi"

 

 

 

 
 
 

 

 

 

 

Q
/(17

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the
thesis entitled

The Experiences of Target and Non-Target

Confronters of Prejudice

presented by

Jennifer S. Pratt-Hyatt

has been accepted towards fulfillment
of the requirements for the

MA. degree in PsychologL

 

 

pm A/ M

Major Professor’s Signature
5/lo /O "7

Date

 

MSU is an aflinnative-action, equal-opportunity employer

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IPrOj/Acc&Prele|RClDateDue,indd

 

THE EXPERIENCES OF TARGET AND NON-TARGET
CONFRONTERS OF PREJUDICE

By

Jennifer S. Pratt-Hyatt

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MATER OF ARTS
Department of Psychology

2007

ABSTRACT

THE EXPERIENCES OF TARGET AND NON-TARGET
CONFRONTERS OF PREJUDICE

By
Jennifer S. Pratt-Hyatt

Researchers have proposed that targets of prejudice engage in a cost-benefits
analysis when deciding whether or not to confront such prejudice. A relatively
overlooked possibility is that non-targets can and do confront prejudice that is directed at
other groups. This study aimed to compare the experiences of target and non-target
confronters of prejudice. University undergraduates engaged in an online interaction,
during which their (confederate) partner made either a rude or sexist comment. They
then wrote an essay (in which they could confront their partner) and completed measures
of affect, meta-perceptions of their partner, and perceptions of essay effectiveness.
Results indicated that there were no differences on dependent measures between
participants who confronted rudeness or sexism, or between male and female confronters
of sexism. Those participants who confionted more directly did seem to incur higher

costs (e. g., affect, partners’ evaluation), but also believed they were more effective.

ACKNOWLEDGEMENTS

The author would like to thank Isis Settles, Debby Kashy, Norbert Kerr, and
Cheryl Kaiser for their assistance in the development of this manuscript. I would also
like to thank Lindsay Dietz for assistance with data collection and the coding of

qualitative data.

iii

TABLE OF CONTENTS

LIST OF TABLES .......................................................................... v
INTRODUCTION ........................................................................... 1
METHOD ................................................................................... 15
RESULTS ................................................................................... 24
DISCUSSION .............................................................................. 44
APPENDICES
A — Directions given to participants for online pilot study 1 ................ 61
B — Comments tested in pilot study 2 .......................................... 62
C — Directions given to participants for pilot study 2 ........................ 63
D —- Individual differences scales administered ............................... 65
E — Suggested questions given to participants in main study ............... 66
F — Participant affect .............................................................. 67
G — Perceptions of participant affect ........................................... 68
H — Perceptions of partner evaluation of participant ......................... 69
I — Perceptions of essay effectiveness ........................................... 70
J — Manipulation and suspicion check .......................................... 71
REFERENCES ............................................................................. 94

iv

LIST OF TABLES
Table l - T-test Comparisons of Rude and Sexist Comment Evaluations From Pilot
Study .................................................................................................. 72

Table 2 - Number of Participants who Confronted and Did Not Confront by Sex and
Condition .......................................................................................................................... 73

Table 3 - Correlations and Inter-item Reliabilities of Dependent Variables —— Whole
Sample ................................................................................................. 74

Table 4 - Correlations and Inter-coder Correlations of Coded Essay
Variables .............................................................................................. 75

Table 5 - Means and Standard Deviations of Participant Affect Variables by Condition,
Sex, and Confronter Status ........................................................................ 76

Table 6 - Means and Standard Deviations of Participant Affect Individual Items by
Condition, Sex, and Confronter Status ........................................................... 77

Table 7 - T-tests of Categorical Coded Variables Among Confronters ..................... 79

Table 8 - Correlations and Descriptive Statistics of Continuous Coded Variables and
Participant Affect Variables Among Confronters .............................................. 80

Table 9 - Means and Standard Deviations of Partner Affect Variables by Condition, Sex,
and Confronter Status .............................................................................. 81

Table 10 - Means and Standard Deviations of Partner Affect Individual Items by
Condition, Sex, and Confronter Status ........................................................... 82

Table 11 - T-tests of Categorical Coded Variables Among Confronters ..................... 84

Table 12 - Correlations and Descriptive Statistics of Continuous Coded Variables and
Partner Affect ........................................................................................ 85

Table 13 - Means and Standard Deviations of Partner Evaluation Variables by Condition,
Sex, and Confronter Status ......................................................................... 86

Table 14 - Means and Standard Deviations of Participant Affect Variables by Condition,
Sex, and Confronter Status ........................................................................ 87

Table 15 - T-tests of Categorical Coded Variables Among Confronters ..................... 89

Table 16 - Correlations and Descriptive Statistics of Continuous Coded Variables and

Partner Evaluation Variables Among Confronters .............................................. 90
Table 17 - Means and Standard Deviations of Effectiveness Variable by Condition, Sex,
and Confronter Status ............................................................................... 91
Table 18 - T-tests of Categorical Coded Variables Among Confronters ..................... 92

Table 19 - Correlations and Descriptive Statistics of Continuous Coded Variables and
Dependent Variables Among Confronters ....................................................... 93

vi

Introduction

It is an often-stated goal of psychological research to reduce prejudice and
discrimination. Accordingly, a great deal of research has been conducted in pursuit of
this goal. Prejudice reduction strategies can be classified into two groups: those aimed at
changing the behavior of others (other-focused) and those aimed at changing one’s own
behavior (self-focused). For example, Matt may come to realize that he holds prejudicial
attitudes towards Black people. If he is motivated to change those attitudes, there are a
number of things he can do. He could try to spend time with Black people in order to
revise his attitudes or he could try to “put himself in the shoes” of a Black person in order
to better understand that person’s experiences. These would be examples of self-focused
strategies in which the individual initiates and engages in behaviors aimed to change him
or herself. In the case of other-focused strategies, the individual engages in behaviors in
an attempt to change the attitudes or behaviors of another person. For example, Eve may
notice that Harry holds prejudicial attitudes towards Black pe0ple when he makes an
inappropriate joke. If Eve is motivated to change Harry’s attitudes, she might confront
him about the joke and explain why it was offensive. In this case Eve is attempting to
modify the attitudes of another person and is thus using an other-focused strategy.

Previous research has typically focused on what targeted individuals and groups
can do to reduce the prejudice they face and what non-targets can do to reduce their own
prejudicial behaviors. Thus, psychologists study the use of other-focused strategies by
targeted group members (e. g., racial minorities, women; Kaiser & Miller, 2004; Swim &
Hyers, 1999) and the use of self-focused strategies by people who are not typically

targets (e. g., Whites, men; Tropp & Pettigrew, 2005; Vescio, Sechrist, & Paolucii, 2003).

Few focus on the use of other-focused strategies by non-targets and the role of non-
targeted individuals as allies or advocates in prejudice reduction in others (Czopp &
Montieth, 2003). However, people who are not traditionally targets of prejudice may
have a unique role to play in the fight against discrimination and prejudice. This paper
will review prejudice-reduction strategies and focus particularly on the experiences of
those who confront discrimination directly.
Self-focused Strategies

Several strategies focus on actions taken by people who want to reduce their own
levels of prejudice. Contact theory is one such theory that has received a great deal of
research attention. Initially proposed by Allport (1954), proponents of this theory believe
that intergroup conflict is best ameliorated by promoting contact between different
groups. Interpersonal contact with outgroup members is thought to increase
understanding thereby reducing prejudice. However, this theory has received mixed
empirical support. There is evidence that having certain forms of contact, such as
positive interactions with outgroup members (especially when group memberships are
salient) or having an outgroup friend, are associated with less prejudicial attitudes
towards that outgroup as a whole (Batson et al., 1997; Marcus-Newhall, Miller, Holtz, &
Brewer, 1993; Wright, Aron, McLaughlin-Volpe, & Ropp, 1997). Further, these effects
can occur when people simply know that one of their ingroup friends has an outgroup
friend (i.e., “A friend of a friend is Latino”). On the other hand, some studies suggest
that because of the anxiety and discomfort contact creates, such interactions often serve
only to increase negative attitudes and strengthen stereotypes (Stephan & Stephan, 1985).

Detractors of the theory also note that because there is such a “grocery list” of conditions

that must be met for contact to have benefits (e. g., identities are salient, equal status of
individuals), the strategy has little practical utility (Wright et al., 1997). Researchers
have documented some of these real-world barriers. For example, Shelton and Richeson
(2005) found that while both Black and White participants reported a desire to interact
more with people of the other race, they believed that others did not want to have
increased contact with them. They attributed their own hesitation to initiate interactions
to fear of rejection, but cited lack of interest as the reason that people of other races failed
to initiate interactions. Such misperceptions typify hindrances to the real-world
effectiveness of contact theory as a means of reducing prejudice.

Another proposed strategy for reducing one’s own prejudice is that of attempting
to identify with a superordinate, inclusive group. It is believed that group relations can
be improved by deemphasizing group identity at the level of conflict (e. g., between racial
groups) and stressing identification at the level of a higher, shared group (e. g.,
nationality). For example, to improve race relations in their schools, administrators
might be encouraged to emphasize a shared school identity (e. g., promote school spirit
and rivalries with other schools). Some researchers describe this strategy as switching
the focus from “us” and “them” to “we” (Dovidio, Gaertner, & Validzic, 1998).
Although a popular theory, there is mounting evidence against it. For example, Homsey
and Hogg (2000) found that efforts to promote superordinate groups while minimizing
subgroups led to feelings of hostility over losing the smaller groups’ uniqueness.

Homey and Hogg (2000) proposed that for this strategy to reduce prejudice
subgroup identities must be maintained in addition to emphasizing subordinate group

membership. They found that participants in the experimental condition in which

participants focused on only a superordinate identity (their university) demonstrated more
bias than did participants who focused on subgroup identity (their academic department).
However, very little intergroup bias occurred when both were emphasized. Similarly,
Smith and Tyler (1996) found that White participants who labeled themselves as
Americans in addition to White were more likely to support Affirmative Action than
those who used only the ethnicity label in the absence of American. Thus there is some
evidence that prejudice can be reduced among people who focus on their memberships in
inclusive groups in addition to traditional subgroups.

Perspective-taking is another potential prejudice-reduction strategy aimed at
reducing one’s own prejudice. It is believed that “putting oneself in someone else’s
shoes” is associated with empathy arousal which in turn leads to more positive intergroup
attitudes. Batson et a1. (1997) manipulated the degree to which study participants
empathized with a stigmatized outgroup member (e. g., a gay man with AIDS).
Compared to the low-empathy condition, participants in the high-empathy condition
showed significantly more favorable attitudes to the stigmatized group. Similarly, in a
study by Vescio et a1. (2003), participants listened to a radio interview in which an
Afi‘ican American man described difficulties and challenges he had faced. Those
participants who were asked to think about the feelings the man experienced as they
listened subsequently reported more favorable attitudes towards African Americans than
participants who were asked to remain objective as they listened. In sum, it appears that
prejudice can be reduced among people who empathize with targets.

The strategies discussed thus far are ways in which individuals’ own behaviors

change their attitudes towards other groups (e.g., through seeking contact with other

groups or taking the perspective of others). Although there is some evidence that they
can be effective, it is often found that this is true only under very specific conditions
making their real-world utility questionable. Further, as self-focused strategies, they
must be carried out by those who hold prejudicial attitudes and wish to change them.
They are of little use when it comes to people who do not realize they are prejudiced or
do not care to change their prejudicial behaviors and attitudes. There is little opportunity
for targets of prejudicial behavior to create change using these strategies. To reduce
prejudice among such people, “other-focused strategies” must be employed.
Other-focused Strategies

Another body of literature has examined ways in which an individual’s behavior
can change another person's attitudes. This literature has focused primarily on the role of
the direct confrontation of acts of discrimination as a way for targeted individuals to try
to change others’ prejudicial behavior. Kaiser and Miller (2004) define confronting
discrimination as a “volitional process aimed at expressing one’s dissatisfaction with
discriminatory treatment to a person or group of people who are responsible for engaging
a discriminatory event” (p.168). Targets of prejudice sometimes choose to confront
those who perpetrate discrimination and a growing body of research has examined the
outcomes associated with such confrontation.

Regarding confrontation, there is an important distinction between situations in
which someone has behaved in a discriminatory way and situations in which someone
has behaved rudely in general. Confrontation in each situation results in different
consequences and these outcomes seem to influence rates of confronting. Research has

shown that target-group confronters incur social costs (e. g., are derogated, are liked less)

when confronting discriminatory behavior. However, they incur costs to a lesser extent
when confronting other types of more general offensive behavior (e. g., rudeness; Kaiser
& Miller, 2003; Shelton & Stewart, 2004). Further, Shelton and Stewart (2004) found
that women in a high social costs condition were less likely than those in a low cost
condition to confront sexist behavior. However, there were no group differences in rates
of confronting general rude behavior. Thus, people seem to consider social costs when
deciding whether or not to confront discrimination, but not when deciding whether or not
to confront rudeness. This research suggests that there are unique costs associated with
confionting discrimination, which will be discussed in greater detail below.
Confrontation by targets

Confrontation may be associated with benefits for both the confronter and her/his
larger social group (Kaiser & Miller, 2004). By bringing attention to acts of
discrimination, one has a chance to educate the perpetrator and perhaps alter her/his
future behavior. This may result in fewer obstacles to the confronter’s personal goal
pursuits (e.g., ending sexual harassment that was interfering with one’s work
performance) as well as helping other potential targets who encounter the former
perpetrator in the future. Some benefits of confronting involve the absence of negative
consequences associated with not confronting. By confronting, targets may avoid the
intrapersonal costs associated with ignoring prejudice. According to self-discrepancy
theory (Higgins, 1987), if a person who believes s/he ought to confront discrimination
ignores a discriminatory comment, s/he is likely to experience guilt or anger at the self
(Shelton, Richeson, Salvatore, & Hill, 2005). In contrast, the absence of such self-critical

emotions could be considered a benefit of confronting.

While there are potential benefits associated with targets confronting prejudice,
many recognized acts of prejudice go unconfronted (Kaiser & Miller, 2004; Swim &
Hyers, 1999; Woodzicka & LaFrance, 2001). For example, Swim and Hyers (1999) had
women participate in a group discussion during which a male confederate made a series
of sexist comments. Just over half (55%) of the women confionted at least one of the
sexist remarks in some way (e. g., questioning the confederate, making a sarcastic
comment, grumbling). Of the 45% who did not confront, 75% later reported that they
found the man’s remarks sexist. Why might these women have chosen not to confront
these sexist remarks? While confronting an act of prejudice may bring about the desired
reduction in prejudicial behavior, this gain often comes at a price to the confronter.

Victims of prejudice often report fears that confrontation will result in negative
perceptions by others and even retaliation. These fears may be justified. Several studies
by Kaiser and Miller (2001 , 2003; see also Garcia, Reser, Amo, Redersdorff, &
Branscombe, 2005) have shown that targets who confront prejudice are evaluated harshly
by observers. For example, when an Afi'ican American student attributes a poor peer
review on an experimental task to prejudice, others see him as a complainer, a
troublemaker, and hypersensitive. This is true even when raters are aware that prejudice
is in fact occurring. Such negative evaluations could have important real world
consequences. Kaiser and Miller (2003) propose that dominant group members retaliate
against claims of discrimination because such claims may threaten their position in the
status hierarchy. Discrimination claimants may have negative interpersonal experiences,
have resources withheld, or be less likely to be hired or promoted. So while targets who

confiont prejudice may benefit from the act they also face considerable risks by doing so.

Confrontation by non-targets

Past research on discrimination confrontation has focused almost exclusively on
target- group member confi'onters. Most studies of discrimination focus on the
experiences of women and People of Color as targets of prejudice and potential
confronters. In such studies, men and Whites are usually the perpetrators of
discrimination. However, people who are traditionally considered perpetrators (e. g.,
White people, men) may have a unique role to play in the fight against prejudice and
discrimination as allies and advocates of discrimination targets. For example, a
heterosexual person may comment on the inappropriateness of a joke about homosexuals
or a White person may work with an organization aimed at promoting civil rights.

While there is a history of non-targets acting as allies and advocates for targeted
groups (e. g., Whites in the Civil Rights Movement, men in the Women’s Movement,
straight people advocating for Gay Rights), little empirical research has addressed their
experiences or effectiveness in these roles. One line of research has begun to study non-
target confronters empirically. Czopp and Monteith (2003) had participants imagine
being confronted about a discriminatory act by either a target (Black person for a racist
act or woman for a sexist act) or non-target (White person for a racist act or man for a
sexist act). They found that people believed they would feel more guilt and experience
less tension when confronted by someone who was not a member of the targeted group
than when confronted by a target-group member. This was true for ingroup and outgroup
members (i.e., men and women reported they would experience more guilt when
confronted by a man about a sexist act). However, subsequent research involving actual

(not hypothetical) interactions have yielded the opposite results. Winslow (2004) found

that White participants were more distressed by accusations of prejudice fi'om a Black
confronter than they were by the same accusations from a White confronter. Similarly,
Czopp and Monteith (2006) found that confrontation from a Black person elicited more
self-directed negative affect than the same confrontation from a White person. However,
Black and White confronters were equally effective in decreasing participant’s
subsequent stereotypic responses. At this point, it remains unclear if non-target and
target confronters of prejudice elicit different emotional responses and behavioral
outcomes in those they confront.

Cost-benefits analysis: The decision to confront

Because target confronters often encounter negative outcomes, a target of
prejudice may weigh potential negative outcomes against potential positive outcomes
when deciding whether or not to confront. Targets of discrimination are thought to
engage in a costs-benefits analysis or cognitive appraisals (Lazarus, 1999; Kaiser &
Miller, 2004). People are less likely to confront if they believe the potential costs (e. g.,
interpersonal, financial) to be greater than the possible benefits (e. g., prejudice reduction,
less personal guilt).

When the potential social costs to targets of prejudice are high (such as in an
interview setting when a job is on the line), they are less likely to confiont than when
costs are low. Studies have shown that people underestimate the role of social costs in
their intentions to confront or not (Shelton & Stewart, 2004; Swim & Hyers, 1999;
Woodzika & LaFrance, 2001). For example, Shelton and Stewart (2004) found that when
asked to predict how they would react if they were asked sexist questions during an

interview, most women said they would confront in both high and low social costs

conditions. In a second study, in which participants actually faced sexist interview
questions, 92% of women in the low social cost condition did indeed confront the
interviewer. However, in the high social cost condition, only 22% confronted. Thus,
people seem to believe they will confiont prejudice regardless of social costs, however
few actually do so when costs are high.

It is proposed that there are also costs and benefits associated with confrontation
by a non-target, and that they too weigh these possibilities when deciding whether or not
to confront prejudice. It is possible that non-targets will face fewer costs than will
targets, making them more likely to confront. If one hopes to encourage non-targets to
confront, it is important to understand what costs and benefits they believe they will incur
when confronting.

Eflectiveness of confrontation

As stated above, confrontation by a target can be effective. This is particularly
true when the perpetrators hold egalitarian attitudes (Czopp & Monteith, 2003).
Individuals differ in the extent to which they endorse egalitarian ideals or believe that all
people should be treated equally (Czopp & Monteith, 2003; Shelton et al., 2005).
Consistent with self-discrepancy theory (Higgins, 1987), people who have personal
standards to behave in non-prejudiced ways experience self-criticism and guilt when they
behave in prejudicial ways (Devine, Monteith, Zwerink, & Elliot, 1991). Over time,
these self-discrepancies and related emotions serve to reduce prejudicial behavior
(Monteith, 1993; Czopp & Monteith, 2006). However, this process cannot begin if
people are unaware that they are behaving prejudicially. Therefore, confronting acts of

prejudice can play an important role in changing the behavior of those motivated to be

10

unprejudiced. The most successful confrontations involve an interaction between the
target’s behavior (confronting) and the perpetrator’s attitudes (e. g., egalitarian).

There are also reasons why confiontation may not be effective when it is carried
out by a target-group member. Research has shown that behavior that is consistent with
social norms is less noticeable than behavior that is not normative (Channouf, Py, &
Somat, 1999). Confrontation of discrimination by a member of a targeted group may be
considered normative and therefore may not receive much notice or consideration.
Similarly, schema-confirming information receives less processing than disconfirrning
(surprising) information (Gigerenzer & Todd, 1999). If perpetrators of prejudice expect
targets to confront discrimination, the information targets convey may be ignored or
processed very superficially by those confronted. Further, targets may be seen as biased
or having a vested interest. Thus, perpetrators may feel that they can disregard what
targets are saying (Petty, Fleming, Priester, & Feinstein, 2001).

There are a number of reasons why people who are not traditionally targets may
have an advantage when confronting prejudice against targeted groups. Many of the
obstacles of the target group member who confronts are not applicable to non-target
group members. For example, confrontation by non-targets is likely less normative and
less consistent with schernas than confrontation by targets. Further, because they are not
directly affected by discrimination, their perspectives may seem more objective (less
biased) and therefore receive more consideration. Social Comparison Theory (Festinger,
1954) offers additional support for the idea that non-targets may be particularly effective
confronters of discrimination. The theory states that when deciding whether one’s own

behavior is appropriate, a person compares her/his attitudes and behaviors to those of

ll

others. The theory further states that when looking for others with whom to compare
themselves, people prefer those who are similar to themselves. Thus, a confrontation by
a non-target may be more effective because the confronter is more similar to the person
whose behavior s/he is trying to change. For example, if a man is trying to decide if it is
appropriate to make a sexist joke, he may consider the behavior of another man a more
relevant comparison than the behavior of a woman.
The current study

Because prejudice reduction is an important goal to many psychologists, theories
as to how this can be accomplished are abundant. However, many of these theories are
self-focused, meaning they rely on those who have prejudicial attitudes to realize their
attitudes are inappropriate, care enough to do something about it, and enact an often
complicated strategy to change their attitudes. Direct confrontation of prejudiced
individuals is one available other-focused strategy and it appears to be effective in some
circumstances. Unfortunately, there are burdens incurred by targets of prejudice who
confront. Research has proposed that targets engage in a cost-benefits analysis when
deciding whether or not to confront. A relatively overlooked possibility is that non-
targets can and do confront prejudice that is directed at other groups, perhaps even more
effectively than prejudice targets. In addition, non-targets may incur fewer negative
consequences as a result of confronting than would targets. If non-targets can be
effective advocates for societal change, it is important to understand the process through
which people become advocates. It is proposed that once non-targets recognize
prejudice, they engage in a costs-benefits analysis just as targets do. Their perceptions of

their effectiveness and their beliefs about the costs and benefits of confronting are likely

12

to be important factors in their decisions to confront on behalf of targeted groups. This
study aims to compare the lived experiences of target and non-target confionters.

In an attempt to understand the experiences of target and non-target confronters of
discrimination, participants were encouraged to confront a male interaction partner who
made either a sexist or rude comment about a female professor during the course of an
online discussion. The study compares the experiences of confionting between female
participants (target confronters) and male participants (non-target confronters). After
confronting, participants were asked to report on the feelings and evaluations they believe
to have elicited in their partner. They were also asked questions about their perceived
effectiveness in the confrontation (i.e., how likely their partner is to reevaluate and
change his behavior). Because previous work has shown that making claims of
discrimination elicits more negative consequences than making claims of mistreatment in
general, a rude comment condition was included as a control so that it could be
determined if experiences with confronting discrimination differ from experiences with
confrontation in general. Additionally, this condition could help determine if observed
differences between men and women are due to target status or to general gender
differences. For example, if women who confront the sexist comment report that they
were less effective than men who confront the sexist comment, in the absence of a control
condition, it would be unclear if the difference was because the women are target group
members or if women are simply less confident, have different response tendencies, etc.
However, if this “gender difference” emerged among sexism confronters but not rudeness
confronters, it would suggest that the difference was not simply a gender difference but

due to a difference in target status.

13

Hypotheses

Participant mood

1a.

1b.

1c.

It was predicted that participants who confronted a rude comment would
experience less negative affect after confronting than would participants who
confronted a sexist comment.

Further, it was predicted that women who confronted a sexist comment would
report more negative affect after confronting than would men who confronted a
sexist comment.

Across conditions, it was predicted that participants who confronted more directly
would experience more negative affect after confronting than participants who

confronted to a lesser degree.

Partner Meta-Perceptions

2a.

2b.

2c.

3a.

Confronted person ’s mood: It was hypothesized that participants who confronted
a sexist comment would expect their partners to experience more negative affect
when confronted than would participants who confronted a rude comment.

It was further predicted that women who confronted a sexist comment would
expect the confronted person to experience more negative affect after being
confronted than would men who confronted a sexist comment.

Across conditions, it was predicted that participants who confronted more directly
would believe their partners experienced more negative affect after reading the
feedback essay than would participants who confronted to a lesser degree.
Confronted person 's evaluation: It was predicted that participants who confronted

a rude comment would subsequently expect the confronted person to evaluate

14

them less negatively (e. g., see them as less hypersensitive, less of a complainer,
like them more) than would participants who confront a sexist comment.
3b. Further, it was predicted that men who confronted a sexist remark would expect
the confronted person to feel less negatively towards them after the confrontation
than would women who confronted a sexist remark.
3c. Across conditions, it was predicted that participants who confronted more directly
would believe their partners evaluated them more negatively after reading the
feedback essay than would participants who confronted to a lesser degree.
Perceived Eflectiveness
4a. Participants who confronted a sexist comment were expected to believe that they
were less effective at influencing their partners’ future behavior than participants
who confronted a rude comment.
4b. Women who confronted the sexist comment were expected to believe they were
less effective at influencing their partners’ future behavior than were the men who
confront the sexist comment.
4c. Across conditions, it was predicted that participants who confronted more directly
would believe they were more effective at influencing their partners’ future
behavior than would participants who confront to a lesser degree.
Method
Pilot Studies
In order to generate the materials needed for the main study, two pilot studies
were first conducted. The goal of this piloting was to find one rude and one sexist

statement which were equal in severity (i.e., one statement that was sexist and another

15

that was matched on numerous qualities but not sexist). Twenty-five undergraduate
students were asked to list ten statements which they could make about a female
university professor and/or her teaching of a course which would be considered sexist
and ten comments which would be considered rude (see Appendix A). The researchers
read these comments and created a list of eleven statements that best represented the kind
of statements needed for the main study. Each student-generated statement was then
paired with a corresponding researcher-generated statement. For example, if a comment
from the student list was sexist, the researchers created a matched statement that was only
rude (see Appendix B). This list was then presented to an additional 50 students who
rated each statement on several factors (e. g., how sexist, rude, hostile it was; see
Appendix C for a complete list). These data were analyzed to find a sexist comment and a
matched comment that was equally rude, but not sexist (see Table 1). All participants in
the pilot studies were recruited through the Psychology Department's Subject Pool and
completed the piloting online. They received one course credit for their participation.
Participants

One-hundred-sixty-four undergraduates (42% female) were recruited from the
Psychology Subject Pool for the main study and received two course credits for their
participation. They participated in experimental sessions with up to three other people.
Participants were randomly assigned to a condition in which they interacted with a
confederate who made either a sexist or rude remark.
Procedure

When signing up for the study online, participants completed several

questionnaires assessing individual differences (see Appendix D for a list of measures);

16

these measures were not used in the current study. At the study session, participants were
greeted by an experimenter and escorted to a private cubicle. They then read and signed
consent forms and read instructions on a computer screen explaining that the researchers
were interested in impression management during online interactions as well as how
people respond to feedback about the impressions they make. They were told that they
would “chat” online with another participant and that they should form an impression of
this person because they would be providing him or her with feedback later in the study.

Each participant then interacted with an ostensibly male confederate (on many
occasions a female research assistant played this role) in a dyadic interaction using MSN
Messenger, an instant messaging (IM) program. The confederate and participants were
given a list of suggested questions (see Appendix E). They were instructed to chat
casually using the provided questions or their own questions. The confederate was
provided with answers to each of the suggested questions and was instructed to answer
any “off-the-cuff” questions in as neutral a way as possible. Any new information
generated in this way was added to the confederate’s “bio” and was used if similar
questions arose in subsequent interactions. Thus the confederate played a consistent
character across interactions.

Given the focus of the study (i.e., reactions to confronting a same-sex or opposite-
sex interaction partner), it was important for participants to know that their partner was
male and that he was aware of the participant’s sex. The participant was informed of the
confederate’s sex when he introduced himself as Kevin. He then asked the participant’s
name. During the course of the interaction, the “male” confederate made either a sexist

or rude comment about a female professor (generated in the pilot study). The confederate

l7

was blind to condition (whether he would make a sexist or rude comment) until it was
time to make the comment. He then drew a slip of paper from a cup that indicated which
comment to send that participant. Before the participant’s response, the interaction was
quickly terminated. The experimenter explained that the computer network had been
“acting up” and instructed participants to begin the next phase of the study.

After the interaction, participants were told that the researchers were interested in
responses to different kinds of feedback. They were told that some people would be
writing different kinds of feedback essays and others would be receiving these essays.
Each participant was then told that she or he would be one of the essay writers and that
she or he had been assigned to the “constructive criticism about impression management”
condition. They were to write about ways in which the other person could have
“performed” better during the interaction. It was stressed that it was very important for
participants to find something to be critical of (even if they themselves did not see any
problems) and that this feedback was actually very helpful to the other person because
s/he could use it to present a more favorable impression in future online interactions.
Participants were given 10 minutes to write their essays, which they then sent to their
partner over the computer. This process was intended to encourage participants to
confront the sexist or rude comment made by the confederate. It was hoped that the other
parts of the interaction would be reasonably innocuous and that there would be little to
criticize except the sexist or rude comment.

Participants then completed questionnaires about their experiences in the study.
Questionnaires were administered using Medialab software. Participants were asked

about their thoughts and feelings about sending the essay to their partner. Upon

l8

completion of the measures, participants were thoroughly debriefed and given an
opportunity to ask questions before being dismissed.
Measures

Participant Affect. Participants’ affect was assessed with a twenty-item adjective
checklist (Watson, Clark, & Tellegen, 1988). Participants were instructed to report their
mood by indicating (with a seven-point scale) the extent to which they felt each adjective
currently described them. Adjectives listed included upset, guilty, angry, proud, nervous,
and afraid (see Appendix F for a complete list). A Participant Positive Affect variable
was created by calculating the mean of participants’ ratings of the following adjectives:
interested, alert, excited, inspired, strong, determined, attentive, enthusiastic, active, and
proud. Similarly, irritable, distressed, ashamed, upset, nervous, guilty, angry, hostile,
jittery, and afraid were used to calculate a Participant Negative Affect variable.
Reliabilities for all measures can be found in Table 3.

Meta-perceptions of partner. Two measures assessed what participants believed
their partner was thinking and feeling. They did so by completing the affect measure
outlined above, only this time as they thought their partner would have filled it out after
reading their feedback essay (see Appendix G). The same items and procedure used to
calculate mean participant affect subscales were used to compute two subscales for mean
perceptions of partner affect: Perceived Partner Positive Affect and Perceived Partner
Negative Affect. Participants were then asked to complete a measure created for the study
which asked them to indicate how they thought their partner felt about them (the
participant) after reading their feedback essay. They rated the extent to which they

thought their partner believed they were being hypersensitive, making a good point, etc.

19

using a seven-point scale (see Appendix H). Factor analysis of these items suggested the
presence of two factors, which were labeled positive evaluation (justified, making a good
point, interesting, self-righteous, aware of important issues, enlightened, being
reasonable, and helpful) and negative evaluation (being hypersensitive, overreacting,
impolite, being unreasonable, a complainer, annoying, emotional, and acting too liberal).
These variables were created by reverse scoring appropriate items and computing the
mean of items within each factor.

Perceived effectiveness. A five-item effectiveness measure adapted from Kaiser
and Miller (2004) assessed the extent to which participants believed the confrontation
changed their partner’s attitudes and future behavior (see Appendix I). An example item
is “To what extent did the essay let the person know that their behavior is unacceptable.”
Items were reverse scored as needed and a mean of all items was calculated.

Manipulation and suspicion check. After completing all measures, participants
answered questions to determine if they noticed the comment made by the confederate
and understood that they were sending him an essay which criticized it (see Appendix J).
They were also asked if anything about the interaction seemed strange or unusual.

Essay coding

Two independent coders read the constructive criticism essays written by
participants. Each essay was coded for the following five variables: whether or not the
rude or sexist comment was mentioned, essay length, essay tone, evaluation of the
comment, and reaction to the comment. Inter-rater reliabilities can be found in Table 4.

An essay was coded as having mentioned the comment if the participant made any

reference to the confederate’s sexist or rude comment (e. g., “Next time you should

20

explain yourself when you say something about how a woman shouldn’t do something,”
“Another reason that [someone] would become annoyed of you, could have been when
you talked about one of your teachers as not having a brain.”). With one exception, all
participants who mentioned the comment expressed at least a mild level of concern or
criticism about it. One participant who praised the comment was excluded from all
analyses. Therefore, all participants who mentioned the comment in the essays were
considered “confronters” of the rude or sexist comment. All essays, regardless of
whether or not the comment was mentioned, were coded for essay length and essay tone.
Essay length was the number of words in the essay and was calculated using the
word count function of Microsoft Word. Essay tone was assessed using three variables:
fiiendly, helpful, and critical. For each variable, the essay was assigned a score between
1 (in no way fiiendly/helpfirl/critical) to 5 (essay was exceptionally
fiiendly/helpful/critical). An essay was given a 1 for fiiendliness if the writer included
g9 pleasantries or “softened” criticisms (e. g., “I enjoyed talking with you” or “While your
responses were quick, I wish you could have elaborated on your answers”). Although
very few essays were viewed as overtly hostile, essays that received a 1 for fiiendliness
tended to be abrupt and aloof. In comparison, essays that received a 3 for friendliness
demonstrated a normative level of politeness (e. g., included some pleasantries or softened
comments). Such essays were civil and polite, but lacked elements of overt friendliness.
Essays that were “as fiiendly as possible” received a 5 for this code. It seemed that the
writers of these essays were making a conscious effort to be fiiendly and outgoing.
These essays represented a tone that one would use when giving constructive criticism to

a close friend.

21

The essay helpfulness codes were determined primarily by the number of
suggestions an essay writer provided. Many writers did not make explicit suggestions
regarding their partners’ future behavior, but did provide information that could be
considered helpful and that could potentially improve their partners’ behavior. For
example, participants often offered insights into how they felt while conversing with their
partners (e.g., “I felt offended by that” or “When he repeated his questions it made me
feel like he wasn’t listening or paying any attention to what I was saying”). Although
such information is not explicit advice, the partner could deduce that he should refrain
from such behavior if he wants to avoid eliciting those feelings in future conversation
partners. Such pieces of information were considered “helpfirl hints.” An essay received
a 1 for the helpfulness code if it provided no hint of a suggestion or advice. Arr essay
received a 3 if it included one complete suggestion (e.g., “Keep in mind about other
peoples’ opinions and feelings about serious matters, which could potentially hurt
someone’s feelings”) or two helpful hints. An essay was given a 5 for this code if it
offered three or more concrete suggestions or six or more hints.

The criticalness code was determined primarily by the number of criticisms in an
essay. Criticisms tended to take two forms: overt (e.g., “I found you boring to talk to and
also hard/awkward to talk to”) and “sugar-coated.” A sugar-coated criticism was one in
which the writer criticized the partner or some aspect of the conversation, but included a
potential justification for the partner (e. g., “You seem like an okay guy but not really
outgoing and a little robotic, maybe that's because of the suggested questions guideline”)
or reflected the problem back to themselves (e. g., “I also think using online lingo is a

little obnoxious, it’s a pet peeve of mine. . .considering that the majority of people use

22

online language, it’s probably just a personal reaction rather than one that can be
generalized as a bad impression”). Although such comments are criticisms, they seem
less critical than overt criticisms. Arr essay was given a 1 on the criticalness code if it
included no form of criticism. An essay received a 3 if it included one overt criticism or
two sugarcoated ones. An essay was given a 5 if it included three overt criticisms or six
sugarcoated ones.

Two additional codes were given to those individuals who “confronted” the
comment (i.e., mentioned the comment in their essay). First, the essay was coded for
how the comment was evaluated. Two codes were possible: 1) neutral (e. g., “I thought
that comment was a little weird. I wasn’t sure what course he was talking about but
maybe it would make more sense if I knew more”) or 2) negative (e.g., “I thought the
comment you made about a man teaching instead of a women is offensive to women”).

Essays that mentioned the comment were also coded for how the participant
reacted to the comment. Four codes were possible (and essays could receive more than
one code if needed): 1) writer expressed uncertainty about comment (e.g., ““1 wish you
hadn’t been kicked off so that I could learn more about that comment so that I don’t make
any rude judgments about you due to that comment”), 2) writer made a suggestion about
behavior (e. g., “His comment about ‘they should hire someone with a brain’ seemed a
little demeaning and I believe he could have simply stated that he didn’t agree with her
teachings instead”), 3) writer labeled the comment (e.g., “I find this statement rather
sexist”), or 4) labeled the person who made the comment (e.g., ‘“‘If he had talked to
another [and] were to have the same conversation with him, probably that person would

[have] been angry with him and probably wrote him off as rude, arrogant, or even

23

sexist”). The participants’ reaction to the comment codes were used in two ways. Each
of the four codes were treated as separate variables (with each being given a presence vs.
absence score); in addition, a continuous variable was created from the four codes in
which the participants were given a score of 1-4 based on their most severe type of
response (i.e., 1 = expressed uncertainty; 2 = made a suggestion; 3 = labeled the
comment; 4 = labeled the person as rude sexist, etc.).
Results

Manipulation check and suspicion

Responses to manipulation check and suspicion items were reviewed. No
participants were excluded due to failure to follow instructions or failed manipulation
checks. Two participants were excluded from all analyses due to extreme suspicion.
They seemed to have deduced the specific hypotheses of the study.
Confrontation rates by condition

Before testing the primary study hypotheses, preliminary analyses were conducted
to determine if confronting differed by sex and condition; the rates of confronting by sex
and condition can be seen in Table 2. A loglinear analysis was conducted to test whether
or not confrontation rates varied across sex and condition. This analysis suggested that
there were differences between the four sex x condition groups [x2 (0, n = 149) = 0.01 , p <
.05]. To further analyze where group differences existed, follow-up analyses were
conducted. Target status was thought to be an important predictor of confronting. If this
is the case, one would expect to find that women in the sexist condition would be more
likely to confront than people in the other 3 conditions (where they were not targets). A

chi—square test of independence of the women in the sexist condition compared to the

24

other three conditions combined suggested that women in the sexist condition were more
likely to confront than people in the other conditions [)8 (1, n = 14.9) = 21.83, p < .05].

To ensure that these results were in fact due to target status, further analyses were
done to test for simple sex and condition effects. Sex, independent of target status, was
not expected to predict confronting. If this was the case, there should be no difference
between rates of confronting among men and women in the rude condition (where target
status is irrelevant). A chi-square test supported this prediction [)8 (1, n = 70) = 2.32, p >
.05]. However, in the sexist condition (where target status is relevant), a significant sex
difference was expected. Specifically, it women (targets) were expected to confront more
than men (non-targets) in the sexist condition. A chi-square test confirmed this [780, n =
79) = 5.83, p < .05]. A simple condition effect, regardless of target status, was expected
to predict confionting. If this was the case, men in the sexist condition would be
expected to confront more than men in the rude condition. A chi-square test supported
this prediction [)8 (1, n = 79) = 6.68, p < .05]. Thus, in predicting confrontation of the
manipulated comment, there was a simple condition effect, in that the sexist comment
elicited more confrontation (across sex). Further, there was a target status effect, in
which women in the sexist condition (targets) confronted more than individuals in the
other three groups (non-targets). However, there was not a simple sex effect. Men and
women confronted equally when target status was irrelevant.

Data analytic approach

A series of 2 (condition: rude comment, sexist comment) x 2 (participant sex:

male, female) x 2 (confrontation status: confronted comment or did not) between-subjects

ANOVAs were conducted to test the main study hypotheses. Significant interactions

25

were then followed-up by simple interaction or simple main effects analyses. When
interactions were marginally significant (i.e., p < .10), the pattern of means is discussed,
but no follow-up tests are reported. In addition, t-tests and correlations between
dependent variables and coded variables from participant essays are presented.
Correlations between study variables and reliability statistics can be seen in Tables 3 and
4.

Hypothesis 1 — Participant aflect

It was predicted that participants who confronted the rude comment would
subsequently report less negative affect and more positive affect than participants who
had confronted the sexist comment (hypothesis 1a). Further, it was predicted that men
who confronted the sexist comment would subsequently report less negative affect and
more positive affect than women who had confronted the sexist comment (hypothesis
1b). Cell means and standard deviations for these analyses can be seen in Table 5.
Analyses did not provide support for these hypotheses. Neither the condition by
confrontation interaction, the sex by confrontation interaction, nor the three factor
interaction were not statistically significant.

However, there was one marginal effect for the interaction between sex and
condition predicting participant negative affect [F ( 1, 141) = 2.87, MSE = 1.46, p < .10].
Women in the rude condition reported higher levels of negative affect (M = 1.83, SD =
0.72) than did men in the rude condition (M = 1.68, SD = 0.71). However, in the sexist
condition, men reported more negative affect (M = 1.81, SD = 0.79) than did women (M

= 1.67, so = 0.65).

26

Additional analyses were conducted to test for group differences on the individual
terms of the affect measures (e. g., angry, proud, guilty; see Table 6). Several significant
and marginally significant effects emerged. There was a main effect of sex predicting
participant’s self-reported level of pride [F (l, 141) = 8.10, MSE = 23.25, p < .05] such
that men reported significantly more pride (M = 3.62, SD = 1.64) than did women (M =
3.09, SD = 1.67). This main effect was qualified by a marginally significant interaction
between sex and confronter status [F (1, 141) = 3.57, MSE = 10.25, p < .10]. Female
confronters reported feeling somewhat less proud (M = 3.00, SD = 1.88) than did female
non-confronters (M = 3.1 1, SD = 1.55). However male confronters reported more pride
(M = 4.12, SD = 1.72) than did male non-confronters (M = 3.45, SD = 1.61).

There was a significant sex effect predicting the participant active variable [F (l ,
141) = 4.28, MSE = 11.22, p < .05]. Men reported feeling more active (M = 3.66, SD =
1.61) than did women (M = 3.43, SD = 1.65). Further, this main effect was qualified by a
significant interaction between sex and confronter status predicting active feelings [F (1,
141) = 5.28, MSE = 13.84, p < .05]. Analysis of simple main effects revealed a marginal
difference between female confronters and non-confronters [F (1 , 141) = 3.00, MSE =
7.87, p < .10]. Women who confronted the comment reported feeling less active (M =
3.03, SD = 1.79) than did women who did not confront (M = 3.69, SD = 1.45). There was
not a significant difference [F (l, 141) = .38, MSE = .99, p > .05] between men who
confronted (M = 3.84, SD = 1.60) and those who did not (M = 3.60, SD = 1.63). In
addition, the interaction between sex and confi'ontation status was qualified by a
marginally significant three-way interaction between sex, condition, and confronter status

predicting the active variable [F (1, 141) = 2.89, MSE = 7.51, p < .10]. Examination of

27

mean trends for this 3-way interaction showed a large difference between the level of
active feelings reported by female confronters of rudeness (M = 1.71, SD = 0.76) and
male confronters of rudeness (M = 4.00, SD = 1.91). All other group means are within a
half of a standard deviation of each other.

There was also a significant main effect of confronter status predicting participant
anger [F (1, 141) = 4.17, MSE = 4.84, p < .05], with confronters reporting more anger (M
= 1.93, SD = 1.34) than non-confronters (M = 1.42, SD = 0.86). In addition, there was a
marginal three-way interaction of sex, condition, and confronter status predicting anger
[F (1, 141) = 2.84, MSE = 3.29, p < .10]. Examination of mean trends suggests a possible
explanation for this interaction. Among confronters, men were angrier than women in
the rude condition and women were angrier than men in the sexist condition. However,
these trends are reversed among non-confronters (see Table 10 for means and standard
deviations).

There were several additional statistically significant and marginally significant
main effects of confronter status predicting individual affect items. Confronters reported
feeling more alert [F (l, 141) = 3.05, MSE = 6.95,p < .10; M= 4.36, SD = 1.43] than did
non-confronters (M = 4.01 , SD = 1.51). Confronters also reported feeling more upset [F
(1, 141) = 9.89, MSE = 11.89,p < .05; M= 2.07, SD = 1.34] than did non-confronters (M
= 1.43, SD = 0.85) and confronters reported feeling more hostile [F (l, 141) = 3.68, MSE
= 3.86, p < .10; M = 1.70, SD = 1.28] than did non-confronters (M = 1.30, SD = 0.76).

To follow up on this trend of participant’s angry affect terms being associated with
confronting, an aggregated “participant anger” variable was created by averaging each

person’s self-report levels of feeling angry, upset, hostile, and irritable (Cronbach’s (it =

28

.82). Group (condition, sex, and confronter status) differences were then tested using the
three-way ANOVA outlined previously. This variable was significantly predicted by
confronter status [F ( l , 141) = 6.84, MSE = 5.72, p < .05], such that confronters reported
higher scores on the aggregated anger variable (M = 1.99, SD = 1.13) than non-
confi'onters (M = 1.52, SD = 0.69). No other main effects or interactions reached
significance.

There was a marginally significant interaction between condition and confronter
status predicting participants’ feelings of interest [F (1, 141) = 3.36, MSE = 5.95, p <
.05]. Confronters in the sexist condition reported feeling more interested (M = 4.38, SD =
1.29) than did rudeness confronters (M = 3.64, SD = 1.28). However, non-confronters in
the sexist condition reported feeling less interested (M = 3.91 , SD = 1.31) than did non—
confronters in the rude condition (M = 4.12, SD = 1.39). There was also an interaction
between condition and confionter status predicting attentiveness [F (1, 141) = 5.84, MSE
= 12.77, p < .05]. Analysis of simple main effects revealed a significant difference
between confronters and non-confronters in the sexist condition [F (1 , 141) = 7.21, MSE
= 15.77, p < .05]. Confronters in the sexist condition reported feeling more attentive (M
= 4.66, SD = 1.46) than did non-confronters (M = 3.93, SD = 1.44). There was not a
significant difference [F (1, 141) = .85, MSE = 1.80, p > .05] between rudeness
confronters (M = 3.93, SD = 1.44) and those in the rude condition who did not confiont
(M = 4.33, SD = 1.46). There was a marginally significant interaction between condition
and confronter status predicting participants’ feelings of jitteriness [F (1, 141) = 2.88,
MSE = 7.49, p < .10]. Confronters in the sexist condition reported feeling more jittery (M

= 2.23, SD = 1.63) than did rudeness confronters (M = 1.57, SD = 0.94). However, non-

29

confronters in the sexist condition reported feeling less jittery (M = 2.33, SD = 1.63) than
did non-confionters in the rude condition (M = 2.48, SD = 1.69).

There was a sex by condition interaction predicting participant distress [F (l, 141)
= 4.55, MSE = 6.01 , p < .05]. Analysis of simple main effects showed a significant sex
difference in the sexist condition [F (1, 141) = 4.16, MSE = 5.50, p < .05]. Women in the
sexist condition reported feeling less distressed (M = 1.68, SD = 0.93) than men in sexist
condition (M = 2.18, SD = 1.33). In the rude condition, there was not a significant
difference [F (1, 141) = .73, MSE = .91,p > .05] between men (M= 1.80, SD = 0.98) and
women (M = 2.03, SD = 1.28). There was also a sex by condition interaction predicting
participant guilt [F (l, 141) = 5.02, MSE = 7.07, p < .05]. Analysis of simple main
effects revealed a significant condition effect among women [F (1, 141) = 5.97, MSE =
8.04, p < .05]. Women in the rude condition reported feeling more guilty (M = 2.10, SD
= 1.51) than women in the sexist condition (M = 1.43, SD = 1.07). There was not a
significant difference [F (1, 141) = .52, MSE = .61, p > .05] between men in the rude
condition (M = 1.51, SD = 1.05) and men in the sexist condition (M = 1.68, SD = 1.11).
Additionally, there was a sex by condition interaction predicting fear [F (1, 141) = 2.97,
MSE = 1.56, p < .10]. Women in the rude condition reported feeling more afraid (M =
1.48, SD = 1.06) than did men in the rude condition (M = 1.34, SD = 0.66). However,
women in the sexist condition reported feeling less afraid (M = 1.18, SD = 0.54) than did
men in the sexist condition (M = 1.36, SD = 0.69).

Across all conditions, it was predicted that participants who confronted more
directly would report less positive affect and more negative affect (hypothesis 1c). Five

of the coded variables were categorical, and participants were assigned to one of two

30

groups for each. The specific categories are as follows: neutral/negative evaluation of the
comment, expressed uncertainty/did not, made a suggestion/did not, labeled the
comment/did not, and labeled the partner/did not. These groups were compared on the
participant affect variables using t-tests (see Table 7). Results indicated that confronters
who evaluated the comment negatively, compared to those who evaluated it in a neutral
or ambivalent way, experienced more negative affect. Participants who expressed
uncertainty about the comment (i.e., said they were not sure what it meant or wanted to
discuss it further), reported more positive affect than confronters who did not express
uncertainty. Compared to participants who did not, participants who labeled the
comment in their essays (e. g., said it was sexist) reported more negative affect.

Correlations were calculated for the participant affect variables and the five
continuous, coded variables assessing confronting directness: fiiendliness, helpfulness,
criticalness, reaction strength (1 — expressed uncertainty to 4 — labeled the partner), and
essay length (see Table 8). Results indicated that participants who reacted to the
comment more severely (e. g., called the person sexist rather than making a suggestion)
reported more negative affect. Essays that were coded as more fiiendly were written by
participants who reported more positive affect.
Hypothesis 2 — Expected Partner Affect

It was predicted that participants who confronted the rude comment would expect

their partners to experience less negative affect and more positive affect than participants
who had confronted the sexist comment (hypothesis 2a). Further, it was predicted that
men who confronted the sexist comment would expect their partners to experience less

negative affect and more positive affect than women who had confi'onted the sexist

31

comment (hypothesis 2b). Cell means and standard deviations for these analyses can be
seen in Table 9. The sex by confrontation interaction and the three factor interaction did
not reach statistical significance, thus not supporting hypothesis 2b. However, there was
a marginally significant effect for the interaction between condition and confronter status
predicting partner positive affect [F (l, 141) = 3.48, MSE = 4.09, p < .10]. Confronters in
the rude condition expected their partners to report less positive affect (M = 3.16, SD =
0.99) than did confronters in the sexist condition (M = 3.66, SD = 1.28). However,
among non-confronters, those in the rude condition expected their partners to report more
positive affect (M = 3.57, SD = 1.20) than did non-confronters (M = 3.37, SD = 1.15).
This pattern of results is counter to hypothesis 2a.

Additional analyses were conducted to test for group differences on the individual
terms of the expected partner affect measures (see Table 10). These analyses revealed a
significant main effect of sex predicting expected partner pride [F (1, 141) = 5.33, MSE =
10.71, p < .05], suggesting that men expected their partners to report more pride (M =
3.26, SD = 1.39) than did women (M = 2.42, SD = 1.39). This effect was qualified by a
significant interaction between sex and condition [F (1, 141) = 5.05, MSE = 10.15, p <
.05]. Analysis of simple main effects revealed a significant sex difference in the sexist
condition [F (1, 141) = 14.71, MSE = 29.56, p < .05]. Men in the sexist condition
expected their partners to feel prouder (M = 3.30, SD = 1.37) than women (M = 2.14, SD
= 1.34). In the rude condition, there was not a significant difference [F (1, 141) = 1.27,
MSE = 2.56, p > .05] between men (M = 3.22, SD = 1.44) and women (M = 2.84, SD =

1.39).

32

There was a significant main effect of participant sex predicting partner guilt [F
(1 , 141) = 4.04, MSE = 8.91, p < .05], with women expecting their partners to feel more
guilty (M = 2.68, SD = 1.64) than men (M = 2.09, SD = 1.42). Confronter status was a
marginally significant predictor of partner guilt [F (l, 141) = 3.09, MSE = 6.81, p < .10],
with confronters expecting their partners to feel more guilty (M = 2.80, SD = 1.72) than
non—confronters (M = 1.98, SD = 1.34). These main effects were qualified by a
marginally significant interaction between sex and confronter status [F (1, 141) = 2.76,
MSE = 6.09, p < .10]. Female confronters thought their partners would feel guiltier (M =
3.20, SD = 1.69) than male confronters (M = 2.20, SD = 1.63). Female non-confronters
also expected their partners to feel guiltier (M = 2.06, SD = 1.45) than male non-
confronters (M = 1.93, SD = 1.29), although this difference was less pronounced among
non-confronters. There was also a significant interaction between condition and
confronter status [F ( 1, 141) = 5.04, MSE = 11.13, p < .05]. Analysis of simple main
effects showed a significant condition effect among confronters [F (1, 141) = 6.30, MSE
= 13.90, p < .05]. Confronters in the sexist condition expected their partners to feel
guiltier (M = 3.06, SD = 1.70) than confronters in the rude condition (M = 1.93, SD =
1.54). Among non-confronters, there was not a significant difference [F (1, 141) = .62,
MSE = 1.15, p > .05] between those in the rude condition (M = 2.05, SD = 1.46) and
those in the sexist condition (M = 1.82, SD = 1.16).

Analyses revealed a marginally significant main effect of sex predicting expected
partner enthusiasm [F (1, 141) = 2.89, MSE = 6.38, p < .10], in which men believed their
partners would feel more enthusiastic after reading their essays (M = 2.98, SD = 1.53)

than women (M = 2.53, SD = 1.43). In addition, there was a significant interaction

33

between condition and confronter status predicting partner enthusiasm [F (1, 141) = 4.97,
MSE = 10.97, p < .05]. Analysis of simple main effects revealed a significant confronter
effect in the rude condition [F (1, 141) = 5.29, MSE = 11.67, p < .05]. Non-confronters
in the rude condition expected their partners to report more enthusiasm (M = 3.02, SD =
1.46) than did confronters (M = 2.00, SD = 1.04). There was not a significant difference
[F (1, 141) = .01, MSE = .01, p > .05] between confronters (M= 2.72, SD = 1.60) and
non-confronters (M = 2.73, SD = 1.57) in the sexist condition.

A marginally significant main effect of confronter status predicted the extent to
which participants thought their partner would feel inspired [F (1, 141) = 3.59, MSE =
7.37, p < .10], with non-confronters expecting their partners to feel more inspired (M =
3.10, SD = 1.47) than confronters (M = 2.52, SD = 1.36). This main effect was qualified
by a marginally significant interaction between condition and confronter status [F (1,
141) = 3.59, MSE = 5.69, p < .10]. Confronters in the sexist condition thought their
partners would be more inspired (M = 2.57, SD = 1.47) than did confronters in the rude
condition (M = 2.36, SD = 0.93). However, among non-confronters, those in the sexist
condition expected less partner inspiration (M = 2.70, SD = 1.55) than did those in the
rude condition (M = 3.34, SD = 1.40).

Participant sex predicted the extent to which they expected their partners to report
feeling active [F (1, 141) = 4.49, MSE = 10.54, p < .05]. Men thought their partners
would feel more active (M= 3.19, SD = 1.51) than women (M= 2.61, SD = 1.51). In
addition, a marginally significant interaction between condition and confronter status
predicted partner active feelings [F ( 1, 141) = 3.62, MSE = 8.51, p < .10]. In the rude

condition, confronters expected their partners to feel less active (M = 2.29, SD = 1.49)

34

than non-confronters (M = 3.26, SD = 1.53). In the sexist condition, confronters expected
their partners to feel more active (M = 2.92, SD = 1.65) than non-confronters (M = 2.88,
SD = 1.52).

There were also several additional main effects predicting the individual items of
the partner affect measure. Men expected their partners to report feeling stronger [F (l ,
141) = 7.91, MSE =15.58,p < .05; M= 3.16, SD = 1.39] than women (M= 2.59, SD =
1.39). Confronters thought their partners would feel more hostile [F (1 , 141) = 4.16,
MSE = 9.52, p < .05; M = 2.75, SD = 1.73] than non-confronters (M = 2.13, SD = 1.35).
Confronters also thought their partners would feel more distressed [F (1, 141) = 3.05,
MSE = 6.52, p < .10; M = 2.93, SD = 1.52] than non-confronters (M= 2.41, SD = 1.36).

Two additional marginally significant interactions predicted partner affect items.
An interaction between condition and confronter status predicted partner feelings of
interest [F (1, 141) = 3.05, MSE = 5.28, p < .10]. Confronters of sexism believed their
partners would be more interested (M = 4.11, SD = 1.31) after reading their essays than
confronters of rudeness (M = 3.50, SD = 1.56). However, non-confronters in the rude
condition thought their partners would be more interested (M = 3.74, SD = 1.33) than
non-confronters in the sexist condition (M = 3.48, SD = 1.30). An interaction between
sex and confronter status predicted the extent to which participants expected their
partners to feel determined [F (1, 141) = 2.83, MSE = 6.51, p < .10]. Female confronters
believed their partners would feel more determined (M = 2.94, SD = 1.80) than female
non-confionters (M = 2.69, SD = 1.33). However, male confronters thought their
partners would feel less determined (M = 2.56, SD = 1.36) than male non-confronters (M

= 3.09, SD = 1.46).

35

It was further predicted that participants (across conditions) who confronted the
comment more directly would expect their partners to report more negative affect and
less positive affect than would participants who had confi'onted less directly (hypothesis
2c). Results indicated that confronters who evaluated the comment negatively, compared
to those who evaluated it in a neutral or ambivalent way, expected their partners to
experience more negative affect (see Table 11). Participants who made suggestions
regarding the comment (e. g., provided ways in which their partner could do better in the
future) expected their partners to experience more negative affect than did participants
who did not make such suggestions. Compared to participants who did not, participants
who labeled the comment in their essays (e. g., said it was sexist) expected their partners
to report more negative affect. Confronters who labeled their partners while discussing
the comment (e.g., called him a sexist person) expected their partners to experience more
negative affect than did participants who did not label their partners.

Analysis of the continuous coded variables indicated that participants who reacted
to the comment more severely expected their partners to experience more negative affect
(see Table 12). Essays that were coded as more friendly were written by participants who
expected their partners to report more positive affect and less negative affect.
Participants who wrote essays that were coded as more critical expected their partners to
experience less positive and more negative affect. Essay length was associated with
expectations for more positive and more negative partner affect. Thus there was some
support for hypothesis 2c. People who were more reactive and critical in their essays
expected less favorable affective outcomes from their partners than did those who wrote

friendlier essays.

36

Hypotheses 3 — Expected Partner Evaluation

It was predicted that participants who confronted the rude comment would expect
their partners to evaluate them more favorably (e. g., see them as more reasonable, less of
a complainer) than participants who confronted the sexist comment (hypothesis 3a).
Further, it was predicted that men who confronted the sexist comment would expect their
partners to evaluate them more favorably than women who had confronted the sexist
comment (hypothesis 3b). Cell means and standard deviations for these analyses can be
seen in Table 13. Analyses did not provide support for hypothesis 3b; both the sex by
confrontation interaction and the three factor interaction were statistically nonsignificant.
However, the condition by confrontation interaction on partner positive evaluation was
statistically significant [F (1, 141) = 4.09, MSE = 3.82, p < .05]. Analyses of simple
main effects showed a marginally significant condition effect among non-confronters [F
(1, 141) = 2.77, MSE = 2.30, p < .10]. Participants in the rude condition who did not
confront their partner expected him to evaluate them more positively (M = 3.67, SD =
0.84) than participants in the sexist condition who did not confront (M = 3.34, SD =
1.03). Although this difference is in the predicted direction (i.e., the rude condition is
associated with perceptions of a more positive evaluation), it only exists among non-
confronters. Among confionters (the group relevant to the hypothesis), the trend is not
statistically significant [F (1, 141) = 1.21, MSE = 1.25, p > .05] and is reversed, with
those in the rude condition expecting a less positive partner evaluation (M = 3.32, SD =
0.92) than those in the sexist condition (M = 3.66, SD = 1.04). Thus these results do not

fully support hypothesis 3a.

37

Additional analyses were conducted to test for group differences on the individual
terms of the evaluation measures (see Table 14). A marginally significant main effect of
confronter status suggested that confronters were more likely to expect their partners to
see them as “acting too liberal” [F (l, 141) = 3.80, MSE = 7.37, p < .10; M = 2.72, SD =
1.61] than non-confronters (M = 2.08, SD = 1.24). This effect was qualified by an
interaction between confronter status and condition [F (1, 141) = 4.80, MSE = 9.30, p <
.05]. Analysis of simple means revealed a significant confronter status effect in the sexist
condition [F (1, 141) = 1 1.37, MSE = 22.05, p < .05]. Confronters of sexism were more
likely to expect their partners to see them as acting too liberal (M = 2.91 , SD = 1.64) than
non-confronters (M = 1.85, SD = 1.06). There was not a significant difference between
confronters (M = 2.07, SD = 1.38) and non-confronters (M = 2.17, SD = 1.31) in the
rudeness condition [F (l, 141) = .07, MSE = .12, p > .05].

Confronter status also predicted expecting to be seen as emotional [F (1, 141) =
3.28, MSE = 6.52,p < .10] and aware ofimportant issues [F (1, 141) = 7.49, MSE =
15.87, p < .05]. Confronters believed their partners would view them as more emotional
(M = 2.85, SD = 1.64), but also more aware of important issues (M = 3.79, SD = 1.65)
than did non-confronters (M = 2.32, SD = 1.29; M = 2.97, SD = 1.26 respectively).

The interaction between confrontation status and condition predicted several
items on the partner evaluation scale. There was a significant interaction predicting the
extent to which participants expected their partners to evaluate them as justified [F (1,
141) = 4.48, MSE = 6.88, p < .05]. Analysis of simple main effects found a statistically
significant difference between confronters and non-confronters in the rude condition [F

(1 , 141) = 5.33, MSE = 5.79, p < .05]. Rudeness non-confronters expected their partners

38

to view them as more justified (M = 3.93, SD = 1.07) than rudeness confronters (M =

3.21 , SD = 0.89). In the sexist condition, there was not a significant difference [F (1,

141) = 0.52, MSE = .97, p > .05] between confronters (M = 3.62, SD = 1.41) and non-
confronters (M = 3.39, SD = 1.36).

In addition, there were several marginally significant interactions. Mean trends
were examined to interpret each interaction. There was a marginal interaction between
condition and confionter status predicting the extent to which participants expected their
partners to evaluate them as interesting [F (1, 141) = 2.84, MSE = 5.51, p < .10].
Confronters of sexism thought their partners would evaluate them as more interesting (M
= 3.49, SD = 1.57) than did confronters of rudeness (2.93, SD = 1.07). Among non-
confronters, those in the in the sexist condition thought their partners would evaluate
them as slightly less interesting (M = 3.33, SD = 1.29) than did rudeness confronters (M =
3.60, SD = 1.34).

There was a marginal interaction between condition and confronter status
predicting the extent to which participants expected their partners to evaluate them as
self-righteous [F (1, 141) = 2.84, MSE = 7.22, p < .10]. Confronters of sexism thought
their partners would evaluate them as more self-righteous (M = 3.51, SD = 1.76) than did
those people who confionted rudeness (M = 2.50, SD = 1.34). Among non-confronters,
those in the in the sexist condition (M = 3.15, SD = 1.46) and those in the rudeness
condition (M = 3.17, SD = 1.57) believed they would be seen as similarly self-righteous.

There was a marginal interaction between condition and confronter status
predicting the extent to which participants expected their partners to evaluate them as

hyper-sensitive. Confronters of sexism thought their partners would evaluate them as

39

more hypersenestive (M = 3.40, SD = 1.53) than did confionters of rudeness (M = 2.79,
SD = 1.67). Among non-confronters, those in the in the sexist condition thought their
partners would evaluate them as less hypersenstive (M = 2.52, SD = 1.37) than did
rudeness confronters (M = 3.16, SD = 1.90).

There was also a marginal interaction between condition and confronter status
predicting the extent to which participants expected their partners to feel they were
making a good point [F (l, 141) = 3.87, MSE = 6.77, p < .10]. Confronters of sexism
were more likely to expect their partners to feel they had made a good point in their essay
(M = 4.30, SD = 1.30) than confronters of rudeness (M = 3.71, SD = 0.99). However,
among non-confronters, those in the rude condition were more likely to expect this
evaluation (M = 4.22, SD = 1.30) than those in the sexist condition (M = 3.89, SD =
1.41 ).

Directness of confrontation was also expected to be associated with the evaluation
participants expected from their partners. Specifically, it was hypothesized that
participants who confronted more directly would expect their partners to evaluate them
less positively and more negatively (hypothesis 3c). Results indicated that, contrary to
predictions, confronters who evaluated the comment negatively, compared to those who
evaluated it in a neutral or ambivalent way, expected their partners to evaluate them more
positively (see Table 15). Compared to participants who did not, participants who
labeled the comment in their essays believed their partners would evaluate them more
positively. Consistent with predictions, confronters who labeled their partners while
discussing the comment expected their partners to evaluate them more negatively than

participants who did not label their partners.

40

Results also indicated that participants who reacted to the comment more severely
expected their partners to evaluate them more negatively (see Table 16). Essays that
were coded as more fiiendly were written by participants who expected their partners to
evaluate them more positively and less negatively than did participants who wrote less
fiiendly essays. Participants who wrote essays that were coded as more critical expected
their partners to evaluate them more negatively. Essay length was associated with more
positive and more negative expected partner evaluations.

Hypotheses 4 — Expected Effectiveness of Essay

It was predicted that participants who confronted the rude comment would
believe they were less effective at changing their partners’ future behavior than
participants who confronted the sexist comment (hypothesis 4a). Further, it was
predicted that women who confronted the sexist comment would believe they are less
effective at changing their partners’ future behavior than men who confronted the sexist
comment (hypothesis 4b). Cell means and standard deviation for these analyses can be
seen in Table 17. The analyses failed to support the hypotheses as the condition by
confrontation interaction, the sex by confrontation interaction, and the three factor
interaction were not statistically significant. However, there were two significant main
effects predicting perceived effectiveness. Across conditions, women believed their
essays would be more effective at changing their partner’s future behavior [F (1 , 141) =
4.91, MSE = 5.86,p < .05; M= 3.38, SD = 1.21], than did men (M= 2.84, SD = 0.99),
Additionally, confronters expected their essays to be more effective [F (1, 141) = 6.59,
MSE = 7.87, p < .05; M = 3.54, SD = 1.21], than did non-confronters (M = 2.82, SD =

1.02).

41

It was predicted that participants who confionted the comments more directly
would expect their essays to be more effective at changing their partners’ future behavior
than would participants who confronted less directly (hypothesis 4c). In support of this
prediction, results indicated that confronters who evaluated the comment negatively,
compared to those who evaluated it in a neutral or ambivalent way, believed their essays
would be more effective at changing their partners’ future behavior (see Table 18). In
addition, those confronters who made a suggestion believed their essays would be
marginally more effective in changing their partner’s future behavior than those who did
not make a suggestion. No other coded variables predicted beliefs about essay
effectiveness (see Tables 18 and 19).

Additional analyses

Analyses were also conducted to examine the effect of condition, sex, confronter
status, and their interactions on the coded essay variables. A series of 2 (condition: rude
comment, sexist comment) x 2 (participant sex: male, female) x 2 (confrontation status:
confionted comment or did not) between-subjects ANOVAs were conducted for the
variables on which all essays received a code, regardless of confronter status (i.e., essay
length, friendliness, helpfulness, and criticalness). These analyses yielded four
statistically significant results. First, confronter status predicted essay length [F (l, 140)
= 7.46, MSE = 23,200.70, p < .05]. The average length of an essay in which the comment
was not mentioned was 122.93 (SD = 52.62) words, compared to 148.85 (SD = 59.85)
words for essays in which the comment was mentioned. There was a condition effect for
the friendliness variable [F (1, 131) = 4.93, MSE = 3.58, p < .05], indicating that

participants in the rude condition wrote fiiendlier essays (M = 2.51, SD = 0.93) than did

42

participants in the sexist comment condition (M = 2.28, SD = 0.77). Participant sex was a
significant predictor of essay helpfulness [F (1, 131) = 8.49, MSE = 8.31, p < .05]. As a
group, women wrote more helpful (M = 2.26, SD = 1.08) essays than men did (M = 1.78,
SD = 0.92). There was a significant interaction between condition and sex, predicting
criticalness [F (1, 131) = 7.74, MSE = 7.14, p < .05]. Analysis of simple main effects
indicated that women in the sexist comment condition were more critical (M = 2.88, SD =
1.06) than were men in the sexist condition [M = 2.08, SD = 0.68; F (1, 131) = 15.34,
MSE = 12.59, p < .05]. In the rude condition, women’s levels of criticalness (M = 2.23,
SD = 0.99) did not differ from men’s [M= 2.39, SD = 1.00, F (1, 131) = 0.41, MSE =
.42, p > .05].

A series of 2 (condition: rude comment, sexist comment) x 2 (participant sex:
male, female) between-subjects ANOVAs were conducted for the variables on which
essays only received a code if the rude or sexist comment was mentioned (i.e., evaluation
of comment, presence of uncertainty, presence of a suggestion, presence of a comment
label, presence of a person label, and reaction to comment). These analyses were only
conducted within the sample of confronters. There was a condition effect on the
comment evaluation variable [F (1, 55) = 4.53, MSE = 1.06, p < .05], indicating that
participants in the sexist comment condition evaluated the comment more negatively (M
= 0.50, SD = 0.51) than did participants in the rude comment condition (M= 0.85, SD =
0.38). There was a condition by sex interaction predicting uncertainty about the
comment [F (1, 56) = 4.32, MSE = .92, p < .05]. Analysis of simple main effects
suggests that there was a significant difference [F (1 , 56) = 5.15, MSE = 1.00, p < .05]

between the uncertainty expressed by men in the rude comment condition (M = 0.01, SD

43

= 0.01) and men in the sexist comment condition (M = 0.44, SD = 0.51). There was not a
significant difference [F (l, 56) = .51, MSE = .11, p > .05] between women in the rude
condition (M = 0.43, SD = 0.53) and women in the sexist condition (M = 0.29, SD =
0.46) on the uncertainty variable41). The sex of the essay writer predicted the presence
of a suggestion in the essay [F (1, 56) = 4.18, MSE = 1.00, p < .05]. Women were more
likely (M = 0.51, SD = 0.51) to make a suggestion than were men (M = 0.28, SD = 0.46).
Condition predicted labeling of the comment [F (1, 56) = 7.51, MSE = 1.64, p < .05];
sexism confronters were more likely to label the comment (M = 0.46, SD = 0.50) than
were rudeness confronters (M = 0.07, SD = 0.27). Condition also predicted the combined
comment reaction variable [F (1 , 56) = 4.28, MSE = 6.72, p < .05]. Confi'onters of
sexism tended to react more directly to the comment (e. g., label the person as opposed to
making a suggestion; M = 2.26, SD = 1.24) than did rudeness confronters (M = 1.43, SD
= 1.28).
Discussion

Numerous studies have demonstrated that confrontation of individuals who
behave prejudicially is one way to reduce discrimination in society (e.g., Czopp &
Monteith, 2006; Monteith, 1993). However, there are burdens incurred by targets of
prejudice who confront (e.g., Garcia, et al., 2005; Kaiser & Miller, 2001; 2003), and they
may engage in a cost-benefits analysis when deciding whether or not to confront
(Lazarus, 1999; Kaiser & Miller, 2004). A relatively overlooked possibility is that non-
targets, people who are not directly affected by a given form of prejudice, can and do
confront prejudice that is directed at other groups, perhaps even more effectively than

prejudice targets. In addition, non-targets may incur fewer negative consequences as a

44

result of confi'onting than do targets. The aim of this study was to compare the
experiences of target and non-target confronters of prejudice; to examine whether non-
target confronters are, in fact, more effective but less negatively personally affected, by
confronting prejudice. The discussion will review the results in the following order —
confronting rude vs. sexist comments, confronting by sex, and directness of
confrontation, addressing the results for all dependent variables (i.e., participant affect,
partner affect, expected evaluation, and expected effectiveness) together in each section.
Confronting rude versus sexist comments

It was hypothesized that, compared to participants who confronted a rude
comment, participants who confronted an interaction partner who had made a sexist
comment would subsequently report: more negative affect, less positive affect,
expectations that their partner would report more negative affect and less positive affect,
expectations that their partners would evaluate them less favorably, and expectations that
their essay would be more effective. Inconsistent with all of these hypotheses, analyses
revealed that there were no differences between participants who confronted the rude
comment and those who confronted the sexist comment on the primary dependent
variables. However, analysis of individual scale items did show that participants who
confronted a sexist comment reported feeling more jittery than participants who
confronted rudeness. Confronters of sexism also expected their partners to feel guiltier
than did participants who confronted a rude comment. Exhibiting sexist attitudes may be
seen as more offensive than exhibiting general rudeness and this difference may have led
sexism confronters to expect their partners to feel guiltier. Although these exploratory

findings could be a Type I errors and may not prove reliable in future studies, they may

45

suggest that confrontation of rude and sexist comments do result in differences in specific
affective outcomes. Perhaps such differences are obscured when they are aggregated
with other irrelevant items in a mood scale.

Previous research has found that people who confront rudeness incur fewer costs
than people who confront prejudice (Kaiser & Miller, 2003; Shelton & Stewart, 2004).
With the exception of the findings regarding jitteriness and partner guilt, the results of the
present study are inconsistent with past research. However, several potentially important
differences exist between previous studies and the current one. First, the format of the
interaction and the confrontation was different in this study than in previous studies. The
current study occurred entirely online and participants never actually met the person
whom they confronted. There may be important contextual differences that influence the
outcomes experienced by confronters. For example, perhaps confronting a person online,
as opposed to in person, leads to less understanding or concern for what the confronted
person is thinking or feeling.

Previous research suggests that costs are a key difference between confrontations
of rudeness and prejudice (Kaiser & Miller, 2003; Shelton & Stewart, 2004). For
example, Shelton and Stewart (2004) found that women were less likely to confront
sexism when potential social costs were high than they were when costs were low.
However, social costs did not affect rates of confronting rudeness. The authors argue that
women fear retribution more when confronting sexism than they do when confronting
rudeness and that in high cost situations they are less likely to risk such retribution. This
suggests that perceived social costs impact how aversive a possible confrontation seems.

There were very few social costs associated with confrontation in this study (participants

46

confronted an equal-status stranger whom they would never meet). Perhaps these very
low social costs in both conditions minimized differences between confronting rudeness
and confronting sexism, leading participants to report similar levels of positive and
negative affect. Participants may have realized that the social costs were low for their
partners as well. If so, they might have concluded that their critical essays would have a
minimal effect on their partner, regardless of whether the essays implied that the partner
was rude or sexist.

Another potentially important distinction between this study and previous ones is
that previous studies have examined confrontation outcomes from the perspective of the
person who was confronted. For example, Kaiser and Miller (2003) asked people who
had been confronted to evaluate the person who confronted them. In contrast, this study
examined outcomes from the confronters’ perspective (their own affect and meta-
perceptions of the confronted person’s reactions). It is possible that people who confront
prejudice are in fact evaluated more harshly than people who confront rudeness, but that
the confronters are unaware of these outcomes.

An additional potential explanation for these findings is that the rude and sexist
conditions may not have differed sufficiently. Although pilot testing suggested that the
rude and sexist comments were seen as similarly rude, but differed in level of sexism, it is
possible that the conditions were not adequately unique from each other to replicate past
findings. In the pilot testing, participants were rating a series of comments, both rude and
sexist. It is likely that ratings of a given comment were influenced by the statements
raters had already read. For example, a sexist comment may have seemed more sexist

when contrasting it with a similar rude comment. Additionally, both the sexist and rude

47

comments in the main study implied that the disliked professor was female. Therefore, it
is possible that participants interpreted the rude comment as somewhat sexist. Unlike
raters in the pilot testing, participants in the rude condition had not just read several
overtly sexist statements to which they could contrast the rude one. However, only one
confronter in the rude condition mentioned anything about sexism in her essay (“His
comment about ‘they should hire someone with a brain’ seemed a little demeaning and I
believe he could have simply stated that he didn’t agree with her teachings instead.
Maybe it was a sexist comment, but I don’t know Kevin, so it might have been
unintentionally harmfirl”), compared to numerous confronters in the sexist condition.
Thus, the comparability of the rude and sexist statements is somewhat unclear.
Male versus female confronters

It was hypothesized that, compared to men, women who confronted an interaction
partner who had made a sexist comment would subsequently report more negative affect,
less positive affect, expectations that their partner would report more negative affect and
less positive affect, expectations that their partners would evaluate them less favorably,
and expectations that their essay would be less effective. Counter to these hypotheses,
analyses of the primary dependent variables revealed that there were no differences
between men and women who confronted the sexist comment (or that any such
differences were not also present in comparisons of men and women in general).

Little previous research has examined the reactions and expectations of people
who confront acts of discrimination. Rather, most research has focused on the reactions
of the person who is confronted. The theoretical basis for this set of hypotheses, that men

and women would experience confronting differently, was all related to the confronted

48

person’s perspective. For example, it was proposed that people pay more attention to acts
that are non-normative or that violate schemas (Channouf et al., 1999; Gigerenzer &
Todd, 1999), suggesting that a confronted person might notice a non-target confrontation
more than one from a target. Social Comparison Theory also suggested (F estinger, 1954)
that a person who wants to determine if she or he is behaving acceptably compares her or
himself to comparable others (i.e., a man who is confronted might find it more personally
meaningful if he is confronted by a man than by a woman). Although these theorized
processes may occur for confronted individuals, the present study examined the
confronter’s perspective.

Even if confrontation from targets and non-targets do influence the confronted
person differently (consistent with theory and some previous studies, e. g., Czopp &
Monteith, 2003), there may not be any reason to expect the confronter to be aware of this.
It is especially likely that the confronters in this study were uncertain about the feelings
and reactions of their partners. They did not have any previous experience with him, they
only communicated online, and they received no feedback from him after the
confrontation. Perhaps participants were not able to determine, or even motivated to
think much about, how their partner reacted. Thus, future studies that examine the
responses of the confronted individual may find the hypothesized relationships, although
the current study of confronters did not.

Several characteristics of the confrontation itself may also explain the failure to
find stronger sex differences. As stated earlier in the discussion of condition effects, the
social costs of this confrontation situation were quite low and this may have made sex

differences unlikely. The lack of any potential retribution may have minimized

49

differences between the experiences of men and women. There may have been little
motivation to process deeply what the other person was thinking or how he might react.
“Confrontation” in this study was relatively low-impact. It is likely that sending an essay
which one was instructed to write, and in which one could balance niceties with criticism,
in the absence of any audience is a very different experience than choosing to verbally
criticizing someone in the presence of others. These weak confiontations may have
lacked the characteristics that would elicit sex differences.

Although there were no sex differences within the sexist condition, there were
some sex differences related to confronting across the rude and sexist conditions. Across
conditions, women who confronted reported feeling less proud than women who did not
confront, whereas for men, confronting was associated with more pride than not
confronting. Although these exploratory findings could be Type I errors and may not
prove reliable in future studies, they may suggest that men and women do experience
different outcomes when confronting. Given that the rude and sexist comments both had
a woman as the target, men may have felt that by confronting the comment they were
acting on behalf of a group to which they do not belong. Such an act could be viewed as
selfless or noble, and thus elicit pride in the men who performed the act. Yet, if these
findings are replicated, it may suggest a means through which non-targets might be
motivated to confront discrimination against others (i.e., focus on the pride one could
take in such an act).

Across conditions, who confronted expected their partners to feel guiltier than
men who confronted and (male and female) non-confronters. This finding is inconsistent

with the hypothesis that men would expect their partners to feel worse. Again, as both

50

comments have a woman as their target, it may be that dero gating an outgroup member (a
woman) to a fellow ingroup member (a man) may be seen as less inappropriate than
expressing that same derogation to a member of the outgroup (a woman). Therefore,
when women confronted such behavior, they were accusing their partner of a more
offensive act than were the men who confronted the same behavior. This may explain
why the women believed their confrontations elicited more guilt in the partner. This
suggests that a derogatory comment might be qualitatively different depending on the
audience to which it is expressed. Additionally, (across conditions) women who
confronted reported feeling more active than men who confronted. Perhaps this reflects
feelings among the women that they were taking action against the derogatory comment.
Given that it is unclear whether the rude comment was perceived as somewhat sexist by
participants, the meaning of the observed sex differences needs further examination in a
less ambiguous experimental context.
Directness of confrontation

It was predicted that, across conditions and sex, people who confronted more
directly would experience more negative affect and less positive affect, would expect
their partners to report less favorable affect, would expect less favorable evaluations from
their partners, and would expect the confrontations to be more effective. Some support
for these hypotheses was found. Generally, results indicated that people who confronted
more directly (e. g., called the person sexist as opposed to making a suggestion) and more
negatively reported more negative affect, expected their partners to experience more
negative affect, expected their partners to evaluate them more negatively, and thought

their essays would be more effective at changing their partners’ future behavior. There

51

were some exceptions (e. g., those who evaluated the comment more negatively expected
their partners to evaluate them more positively) and differences among the specific
coding variables. Yet the overall pattern of results suggest that people who confront
more directly, compared to those who confront to a lesser degree, feel worse, expect their
partners to feel worse and to evaluate them less favorably, but at the same time the more
direct confronters expect better outcomes (i.e., think they are more effective). Thus, it
seems that more direct confrontations are associated with both higher costs (e.g., more
negative affect, diminished relationship with partner) and greater benefits (e.g., making a
difference). These findings are consistent with analyses that found that compared to non-
confronters (across conditions), confronters expected their partners to experience more
negative affect and experienced more feelings of anger themselves. However, they also
thought their essays would be more effective in changing their partner’s future behavior.
Thus it appears that confrontations, especially those that are more negative and direct, are
associated with perceptions of both costs and benefits.

Although these results are consistent with hypotheses, it is important to note that
it is not possible to conclude that the directness of a participant’s essay caused the
reported affect or partner meta-perceptions. For example, it is possible that people who
were already in a bad mood or who believed their partner did not like them were more
likely to confront more directly. Nevertheless, future work on this topic should certainly
take into account the directness of a confrontation when examining confronters’
experiences. Such studies could lead to a more thorough understanding of the process in
which potential confronters weigh costs and benefits associated with confronting, as well

as the outcomes associated with such confrontations. Perhaps people who confront more

52

directly realize that they have made a direct statement and therefore feel more
efficacious. Maybe direct confronters feel more strongly about what they are saying and
believe their audience will perceive such conviction and find it compelling. Identifying
individual differences that predict direct confronting may prove useful in terms of
predicting who will be an effective confronter.

Results also indicated that the tone in which an essay was written was associated
with many of the dependent variables. Individuals who wrote fiiendlier essays reported
more positive affect, expected their partners to report more favorable affect and to
evaluate them better. In contrast, people who were more critical in their essays expected
their partners to feel less favorable affect and to evaluate them more negatively. This
may suggest that participants were aware of the tone of their essay and realized how it
might affect their partners. It may also be that tone is a means through which one can
express approval or disapproval of a partner’s actions without being explicit about these
attitudes. Thus tone should be considered an important characteristic of confrontations in
future work in this area.

Essay length was related to both participant and perceived partner affect.
Participants who wrote longer essays reported more positive affect and more negative
affect. They also expected their partners to report more positive and more negative affect.
Thus, it seems that writing more was associated with feeling stronger levels of affect,
both positive and negative. This could be indicative of pre-existing levels of affect (i.e.,
participants who were already in a particularly positive or negative mood wrote more) or
of affect that was generated through the act of writing the essay (i.e., writing a long essay

which was positive or negative affected one’s mood positively or negatively).

53

Additional Findings

Some exploratory analyses in this study shed light on the factors associated with
choosing to confront a rude or sexist comment. A loglinear analysis showed that people
in the sexist comment condition were more likely to confront than those in the rude
condition, suggesting that making a sexist remark may be seen as more offensive and
therefore more likely to elicit confrontation. Further, women in the sexist condition were
more likely to confront than men in the sexist condition and men and women in the rude
condition. Thus it seems that being the target of a discriminatory comment is associated
with increased likelihood that one will confront it. Additionally, participants’ feelings of
anger, hostility, and upset were related to higher levels of confronting. Given the nature
of this study it is impossible to be certain the direction of this relationship. It could be
that participants who became angrier while conversing with their partners were more
likely to confront. However, it is unclear if the conversation was a source of anger for
some participants, if they were angered by factors unrelated to the study (e. g., personality
dispositions, situational life factors). Further, it is possible that the act of confronting
itself increased participants’ level of anger; perhaps reliving the potentially unpleasant
encounter while confronting it in the essay (as opposed to disregarding or ignoring it) led
these participants to feel angrier than those who did not confront.

Exploratory analyses in this study also identified some of the potential costs and
benefits associated with confronting and not confronting. Compared to participants who
did not confront, confronters expected their partners to experience more hostility and
more distress. Perhaps these feelings were thought to be caused by the confrontation or

maybe participants saw these feelings as general characteristics of a person who makes

54

prejudicial comments. They were also more likely to expect their partners to evaluate
them as too liberal and emotional. On the other hand, confronters also believed that their
essays would be more effective at changing their partner’s future behavior and that their
partners thought they were aware of important issues. Compared to confronters of
rudeness, sexism confronters thought their partners would feel guiltier. Sexism
confronters also expected their partners to evaluate them as more self-righteous,
hypersensitive, and too liberal than rudeness confronters. However, they were also more
likely to believe that their partners thought they were interesting and making a good
point. This pattern of results suggests that confronting, and confronting sexism
especially, is associated with perceived costs (e. g., being evaluated negatively by others),
but also perceived benefits (e. g., being viewed as aware of important issues, feeling
efficacious). However, it is important to note that some of these findings are only
marginally statistically significant and that future studies will need to verify their
reliability. Also, it is not possible to say that the act of confronting led to these feelings
and expectations. It is possible that the view participants had already formed of their
partners influenced their decision to confront. For example, if, during the course of their
conversation, a person felt that her partner thought she was emotional and self-righteous,
she may have perceived confronting as less costly because her partner already disliked
her.

A limitation of this study is that it does provide direct evidence of the costs and
benefits people associate with a potential confrontation. Although the above findings
suggest that those who chose to confront believed there were both costs and benefits

associated with confronting, they do not offer insight into the perceptions of those who

55

chose not to confront. It is unknown whether non-confronters in this study perceived the
costs and benefits of a confrontation in the same way as confronters. Perhaps confronters
perceived fewer costs associated with confronting or that they weighted such costs
differently than non-confronters. It may be that non-confronters assessed a potential
confi'ontation in a similar way as confronters, but chose to not confront in order to avoid
perceived costs. For example, perhaps the costs and risks were assessed similarly by
different people, but they were acted on differently due to personality traits (e.g., high
harm avoidance). Future studies should attempt to understand the thought processes of
both people who ultimately confront and those who do not.

As alluded to above, these data also do not provide clear information about the net
weight of costs versus benefits. For example, how do people determine what level of
benefits justify a given level of costs? And if this balance predicts confronting, is it that
non-confronters and confronters perceive individual costs and benefits differently (e. g.,
one person feels that being considered hypersensitive is more costly than another person)
or that they have similar understandings of the costs and benefits but balance them
differently (e.g., one person needs to perceive a higher ratio of costs to benefits before
s/he confronts)?

The measures of costs and benefits in this study are evaluated entirely from the
potential confronter’s perspective. The data do not speak to actual feelings elicited in
one’s partner or objective levels of essay effectiveness. It has been argued here that a
person’s perceptions of costs and benefits are more relevant predictors of confronting
behavior than actual costs and benefits. However, it will be important for researchers to

further examine real, rather than perceived, outcomes. It would also be interesting to

56

study causes for and outcomes related to discrepancies between real and imagined
consequences of confronting.
Limitations and Future Directions

Although, overall, the findings of this study failed to support the idea that target
and non-target confronters of prejudice experience different outcomes as a result of
confronting, several potential limitations of the study may provide an explanation for the
lack of support. Therefore, it is important for firture studies to examine these possibilities
before concluding that their target status is unrelated to outcomes relating to
confrontation.

First, the experimental situation was quite different from real-life situations in
which people actually confront. Participants were communicating with a stranger online
and were then encouraged to confront what he had said, in the absence of any real
consequences. Also, the exchange was only between the participant and the person s/he
confronted, as opposed to the confrontation taking place in the presence of others. Any
one of these factors could have had important effects on the experiences of participants.
For example, if one is considering confronting a fiiend, he or she must weigh the
potential benefits of such a confrontation against potential damage to an existing
relationship. Such considerations were not relevant in the current study. Future studies
will need to explore people’s reactions in a variety of situations, especially those which
mirror the circumstances under which people confiont in everyday life.

This study only examined the reactions of the person who confronted. Previous
studies have examined the reactions of the person confronted (e. g., Czopp & Monteith,

2003; Kaiser & Miller, 2003). A study in which the perspectives of both parties are

57

investigated would be ideal. This would allow researchers to explore the differences
between the confronted person’s actual feelings and beliefs and the confronter’s meta-
perceptions. Similarly, future studies could attempt to objectively evaluate the
effectiveness of a confrontation (e. g., the extent to which it leads to fewer prejudicial
comments in a subsequent interaction) and compare it to the confronter’s beliefs.

Further, it will be important to study outcomes associated with confrontation of
various forms of discrimination. Different groups of people experience prejudice and
discrimination in society (e.g., racial minorities, homosexuals, people with disabilities)
and perhaps the target group is related to different outcomes. For example, it has been
shown empirically that acts of sexism are evaluated as more acceptable than acts of
racism (Czopp & Monteith, 2003). Therefore, outcomes related to confronting may
differ accordingly, with less confrontation of sexism than racism or other less acceptable
forms of prejudice.

It is also important to note that many participants did not choose to confront the
comments, despite being given an opportunity to do so. Past research suggests that
individuals differ in the extent to which they recognize instances of prejudice and
discrimination (e. g., Operario & Fiske, 2001). There are also individual differences that,
even when one notices discrimination, may influence the likelihood that he or she will
confront. For example, differences in impression management concerns, personality
traits, and self-efficacy might make people more or less willing to confront a comment
they find inappropriate. Further, these individual differences might affect targets and
non-targets differently when they are faced with discriminatory comments. For example,

if a target and a non-target are equally offended by a prejudicially comment, the target

58

may be especially likely to respond because the comment is personally meaningful.
Although not the primary focus of this study, many of these questions can be addressed
with additional data collected from participants in this study. Relevant analyses are
ongoing and any future findings will be reported in subsequent work by the author.

More broadly, it will be interesting to investigate forms of activism related to
discrimination other than confronting prejudicial statements. For example, what makes
targets and non-targets participate in political rallies or donate money to organizations
aimed at reducing discrimination? Non-targets engage in this kind of behavior and future
research should explore their motivations and experiences. It may be that personality
predicts such behavior (e. g., empathy, conscientiousness). Allport’s (1954) theory
suggests that the amount of social contact one has with outgroup members influences
attitudes towards that group and perhaps this translates into activism on behalf of the
group. Further, it has been suggested that one’s own experiences with discrimination (as
a target) increases awareness of discrimination in the future (Operario & Fiske, 2001;
Sellers & Shelton, 2003). Perhaps this salience extends to all forms of discrimination
(even it does not target one’s self).

The current study failed to find direct evidence that confronting rudeness was
perceived as more costly or more effective than confronting sexism. Similarly, women
who confronted sexism did not seem to fare any worse or feel less effective than did men
who confronted sexism. However, secondary analyses did indicate that condition and sex
differences might exist if specific emotions are studied (e. g., guilt and pride). The study
also supported the idea that more direct confrontations are perceived as more effective,

but are also associated with greater costs. Given the potentially significant implications

59

of this line of research, it will be important to fully investigate these questions and to
improve our understanding of the outcomes experienced by target and non-target
confionters. The answers may help those who want to bring about societal change better

target the advocates who will reap the most rewards while incurring the fewest costs.

60

Appendix A

Directions given to participants for online pilot study 1

The researchers conducting this study would like to know what kinds of
comments students consider to be sexist and what kinds of comments they consider to be
rude. To accomplish this, we would like you to list 10 sexist statements and 10 rude
statements. There are no right or wrong responses for this task. We are simply interested
in your honest opinions.

First, please list 10 things that you could say about a female professor and/or her
teaching of a course that you would consider sexist or offensive to women. The
comments should not be about a specific person or course, but just about “a female
professor.” An example might be, “This course would be a lot better if a man was
teaching it.” Remember, these comments do not have to be things that you would
actually say, think, or agree with. We are simply interested in learning what sorts of
comments students would consider sexist.

Space to list comments.

Now, please list ten things you could say about a professor (either male or female)
that you would consider rude in general. Unlike the last set of comments you listed, these
statements should not be sexist or offensive to women in particular, but rather just rude or
offensive in general. For example, “That professor is such an idiot.” Again, these do not
have to be actual statements that you would choose to make or agree with.

Space to list comments.

61

Appendix B

Comments tested in pilot study 2

 

SEXIST COMMENTS

RUDE COMMENTS

 

10.

ll.

. I don’t know who hired her to teach

the course. This is something that
should be taught by a man.

She doesn’t know anything. She
probably has her degree in baking
cookies.

I bet she only got hired here because
she slept with one of her bosses.

I don’t know why she’s even
teaching. She should be at home
taking care of her family.

She’s a hard ass about assignments —
female professors are so anal
retentive.

She should be teaching a preschool
class where she belongs instead of an
actual college class.

Where did this woman get her
degree? Beauty school?

I always get distracted during her
lectures. I would pay more attention
to what she was saying if she didn’t
wear such slutty outfits.

Everyone knows that male professors
are more qualified than female
professors.

She probably only got a job here
because they needed to fill some sort
of quota for women.

I think we’d be better off if they
didn’t let women be professors.

10.

ll.

. I don’t know who hired her to teach

the course. This is something that
should be taught by a person with a
brain.

She doesn’t know anything. She
probably has a degree in basket
weaving.

I bet she only got hired here because
she knows someone important.

I don’t know why she’s even
teaching. She should be doing
something she’s at least somewhat
competent at.

She’s a hard ass about assignments —
professors are so anal retentive.

She should be teaching a class she
actually knows something about.

Where did this professor get her
degree? Clown school?

I always get distracted during her
lectures. I would pay more attention
to what she was saying if she didn’t
wear such ridiculous outfits.

Everyone knows that some professors
are more qualified than others and this
woman is totally unqualified.

She probably only got a job here
because the department was
desperate.

I think we’d be better off if they
didn’t let morons be professors.

 

62

Appendix C

Directions given to participants for online pilot study 2

The researchers conducting this study would like to know how students evaluate
various comments that someone could make about a university professor. Please read
each statement listed and share your thoughts about it by answering the questions which
follow. There are no right or wrong answers to these questions. We are simply interested
in your honest opinions.

Each statement will be listed individually and followed by these questions.
“1 — not at all” to “7 —— extremely”

How sexist or offensive to women does this comment seem to you?

How sexist or offensive to women do you think the speaker intended this comment to
be?
How rude or impolite does this comment seem to you?

How rude or impolite do you think the speaker intended this comment to be?
How hostile does this comment seem to you?

How hostile do you think the speaker intended this comment to be?

How socially appropriate does this comment seem to you?

How socially appropriate do you think the speaker intended this comment to be?
How rational is this comment?

How rational do you think the speaker intended this comment to be?

How funny does this comment seem to you?

How funny do you think the speaker intended this comment to be?

How politically incorrect does this comment seem to you?

How politically incorrect do you think the speaker intended this comment to be?
How offensive does this comment seem to you?

How offensive do you think the speaker intended this comment to be?

How likely is it that a typical MSU student would make this comment when talking
with other students? (“1 — not at all likely” to “7 - extremely likely”)

To what extent do you think this is an appropriate or acceptable comment to make to
another student? (“1 — totally inappropriate” to “7 — totally appropriate)

How would you feel if someone made this comment about you? (angry, sad, hurt,
amused, confused, happy, proud, ashamed, other )

63

Imagine that you are talking to an acquaintance and he or she makes this comment.
How likely are you to respond in a way that indicates you disapprove of the comment?
(“1 — not at all likely” to “7 — extremely likely”)

What were the speaker’s intentions in making this comment? (open-ended)

If another student said this to you, what would you infer about his or her character?
(open-ended)

Do you have any other comments about this statement (regarding its content,

understandability, etc.) or how you might react or respond if someone said it during a
conversation? (open-ended)

After all statements are evaluated:
Please provide us with a little bit of information about yourself.
What is your gender? (male, female)

In general, to what extent do you think sexism or discrimination against women is a
problem in society? (“1 — not at all a problem” to “7 — an extremely serious problem”)

To what extent are you personally concerned about sexism or discrimination against
women in society? (“1 — not at all” to “7 — extremely”)

In general, to what extent do you think rudeness or impolite behavior is a problem in
society? (“1 — not at all a problem” to “7 — an extremely serious problem”)

To what extent are you personally concerned about rudeness or impolite behavior in
society? (“1 — not at all” to “7 — extremely”)

64

00

Appendix D

Individual differences scales administered
. Demographics
The Mini-IPIP Scales (Donnellan, Oswald, Baird, Lucas, 2006)
Impression management (Leary & Kowalski, 1990)

Gender centrality and regard (modified version of Seller et al.’s, 1997
Multidimensional Inventory of Black Identity)

Appendix N - Internal and External Motivation to
Respond Without Sexism Scale (Klonis, Plant, & Devine, 2005)

The Modern and Old-Fashioned Sexism Scale (Swim, Aikin, Hall, & Hunter,
1995)

. Efficacy about confronting (Kaiser & Miller, 2004)

65

Appendix E

Suggested questions given to participants in main study
What’s your name?
What year are you?
What’s your major?
Do you live on campus or off?
Where are you from originally?
What’s your family like?
Do you play any sports?
What do you do for fun?
What’s your favorite movie?
Do you play video games?
What are you planning to do when you graduate?

What’s your favorite type of food?

66

Appendix F

Participant affect

To what extent do you feel each of the following emotions right now?
(“1 - not at all,” “4 — moderately,” “7 —extremely”)

__ interested
__ distressed
__ excited
__ upset

_ strong
_ guilty

__ angry

__ hostile
__ enthusiastic
__ proud

irritable
alert
ashamed

inspired

nervous

determined

attentive

67

jittery
active
afi'aid

Appendix G

Perceptions of partner affect

While reading my essay, my interaction partner will feel:
(“1 - not at all,” “4 — moderately,” “7 —extremely”)

__ interested
__ distressed
_ excited
__ upset

__ strong
__ guilty

__ angry

__ hostile
_ enthusiastic
__ proud

irritable

alert
ashamed

inspired

nervous
determined

attentive

jittery

active

68

afraid

Appendix H

Perceptions of partner evaluation of participant

(“1 - not at all,” “4 — moderately,” “7 —extremely”)

_ being hypersensitive _ being unreasonable
_ overreacting __ a complainer

_ justified __ enlightened

__ making a good point __ annoying

__ interesting __ being reasonable
_ self-righteous __ helpful

__ impolite __ emotional

_ aware of important issues __ acting too liberal

Briefly describe what you think the other participant will think about you.

69

Appendix I

Perceptions of essay effectiveness

(1 “not at all” to 7 “very much so”)

1.

To what extent do you think your essay will cause the other participant to become
aware of or recognize that aspects of his or her behavior were inappropriate?

To what extent do you think your essay will make the other participant behave in a
less inappropriate manner in the future?

To what extent do you think your essay let the person know that their behavior is
unacceptable?

To what extent do you think your essay caused your partner to reconsider their
behavior?

How likely is your partner to behave in a similar way in the future? (l-very
unlikely, 7-very likely)

70

Appendix J

Manipulation and suspicion check

Please briefly describe your memories of your partner’s behavior during the online
interaction (e. g., what things s/he said, how s/he acted).

Please briefly describe the content of the essay that you sent your partner.
Did anything about this study seem strange or unusual to you?
If yes, at one point in the study did you notice these things?

During the online interaction, did your partner make any comments that could be
considered offensive (by you or others)? If so, please recount what s/he said.

71

Table 1

T '-test Comparisons of Rude and Sexist Comment Evaluations from Pilot Study

 

 

 

Rude Comment Sexist Comment
t
M SD M SD
(df)
How sexist or offensive to _ 6 30*
women does this comment 5.78 1.54 3.33 2.18 (39')
seem to you?
How rude or impolite does -0.09
this comment seem to you? 5'80 1'24 5'83 1'39 (39)
How hostile does thrs 4.3 .1 1.61 4.64 2.01 -0.99
comment seem to you? (38)
How socially appropriate does 1.42
this comment seem to you? 3'29 1'83 2'63 1'96 (37)
How politically incorrect does -3.55*
this comment seem to you? 4'24 2'06 5'58 1'64 (37)
. . _ a:
How offensrve does thrs 4.53 2.02 5.45 1.80 2.09
comment seem to you? (3 7)

 

Note: Rude comment was “I don’t know who hired her to teach the course. This is
something that should be taught by a person with a brain”; sexist comment was “I don’t
know who hired her to teach the course. This is something that should be taught by a
man.”

* p < .05

72

Table 2

Number of Participants who Confronted and Did Not Confront by Sex and Condition

 

 

 

Rude Sexist
Confionted Dld not Confronted Dld not
confront confront
Men 7 (18%) 32 (82%) 18 (45%) 22 (55%)
Women 7 (23%) 24 (77%) 28 (72%) 11 (28%)

 

73

Table 3

Correlations and Inter-item Reliabilities of Dependent Variables — Whole Sample

 

 

1 2 3 4 5 6 7
Measures
1. Participant PA (.92)
2. Participant NA .17* (.81 )
3. Partner PA .72* .16* (.91 )
4. Partner NA .09 .62* .12 (.92 )
5. Partner pos. eval. .50"‘ .15 .66* .09 (.82 )
6. Partner neg. eval. .01 .41* .00 .69* -.05 (.91)
7. Effectiveness .17* .21* .08 .35* .26* .21* (.83)

 

Note: PA = positive affect, NA = negative affect, pos. eval. = positive evaluation, neg.
eval. = negative evaluation
* p < .05

74

.320 A

0.03.3385 EB. ~=wa-qoo.m~. 03852.05. godmmfi MES. $56qu

 

 

_ N w A 3 o a m 3 3

_. m<mEmno= 2.08

38339:
N. mxwammg J: 33v

cane—43:3.
3. gamma .3... .3... 2.2:

mammmmmo:
A. having .33... .3... .3... 2.2:

8383"
3. race—ma .3 33 .3... .3 Q8

6030:
a. Wamoaosflo .3... .3... .3... he... .33... A38

8332:
.V. manna—58m 33... .3 33 as J: 3: A33
w. magma—samm .3... 33 .3... .3 .3... new... 32 A38
3. 0138503 .30... .oo .3... .3 .3... .3... )3... .3... A38
39.8%: .3 .3 :2... .3... .3... .3... .Na... .3... .3... l-

 

239. 0:? $33 93 Bosmoaoa :8 Ban 3 maxi" 8389: 8820a ooaom wee Elev—om To.
*
.6 A .3

75

Table 5

Means and Standard Deviations of Participant Aflect Variables by Condition, Sex, and
Confronter Status

 

 

 

 

Confronted Did not confront
Rude Sexist Rude Sexist
M W M W M W M W
PA 3.51 2.80 3.77 3.59 3.43 3.66 3.41 3.29
(1.13) (0.73) (1.06) (1.44) (1.25) (1.11) (1.29) (0.87)
NA 1.73 1.89 1.89 1.79 1.61 1.82 1.80 1.34

(0.59) (0.85) (0.91) (0.74) (0.64) (0.69) (0.74) (0.25)

 

76

Table 6

Means and Standard Deviations of Participant Aflect Individual Items by Condition, Sex,
and Confronter Status

 

 

 

 

Confronted Did not confront
Rude Sexist Rude Sexist
M W M W M W M W
Participant Positive Affect Items
Interested 3.43 3.86 4.44 4.39 3.81 4.46 3.77 4.18
(1.40) (1.21) (1.15) (1.40) (1.55) (1.06) (1.54) (0.60)
Excited 2.43 2.14 3.06 2.75 2.44 3.25 2.82 2.36
(1.27) (1.46) (1.70) (1.60) (1 .46) (1.48) (1.40) (1 .29)
Stron 3.14 3.14 3.94 3.86 3.34 3.29 3.55 3.45
g (1.57) (1.21) (1.89) (1.63) (1.64) (1.63) (1.57) (1.04)
Enthusiastic 3.14 2.14 3.17 2.93 3.03 3.54 3.45 3.18
(1.07) (0.90) (1.98) (1.78) (1.62) (1.86) (1 .60) (1.40)
Proud 4.43 2.14 4.00 3.21 3.47 3.17 3.41 3.09
(1.40) (1.07) (1.85) (1.99) (1.72) (1.74) (1.50) (1.14)
Alert 4.29 4.71 4.72 4.07 4.06 4.17 3.82 3.73
(0.95) (0.76) (1.27) (1.72) (1.68) (1.55) (1 .44) (1.35)
Ins ired 3.29 2.14 2.56 3.29 3.03 3.38 2.73 2.91
p (1.89) (1.46) (1.62) (2.09) (1.49) (1.61) (1.86) (1.51)
Determined 3.29 1.86 3.28 3.46 3.31 3.21 3.00 3.09
(1.80) (0.69) (1.71) (2.06) (1.60) (1.47) (1.48) (1.70)
Attentive 3.71 4.14 4.78 4.57 4.19 4.38 3.95 3.36
(1.89) (0.90) (1.22) (1.64) (1.71) (1.01) (1.50) (1.50)
Active 4.00 1.71 3.78 3.36 3.56 3.75 3.59 3.54
(1.91) (0.76) (1.52) (1.83) (1.64) (1.67) (1.65) (0.93)

 

77

Table 6 (cont’d)

 

 

 

 

Confronted Did not confiont
Rude Sexist Rude Sexist
M W M W M W M W
Participant Negative Affect Items
Distressed 2.00 2.57 2.22 1.79 1.69 1.86 2.14 1.45
(1.15) (1.40) (1.44) (1.03) (0.90) (1.23) (1.36) (0.69)
U set 1.86 2.14 2.06 2.11 1.50 1.25 1.50 1.27
p (0.90) (1.35) (1.47) (1.42) (0.84) (0.53) (1.19) (0.65)
Guilt 1.57 2.29 1.72 1.54 1.41 2.04 1.68 1.18
y (0.79) (1.89) (1.13) (1.29) (0.95) (1.43) (1.21) (0.40)
An r 1.86 1.57 1.67 2.14 1.28 1.50 1.59 1.18
g y (1.21) (0.79) (1.08) (1.58) (0.63) (1.02) (1.10) (0.40)
Hostile 1.29 1.71 1.83 1.71 1.31 1.46 1.27 1.00
(0.49) (1.50) (1.25) (1.44) (0.69) (0.93) (0.88) (0.00)
Irritable 2.29 2.43 2.22 2.29 2.03 1.89 1.82 2.00
(1.11) (1.27) (1.56) (1.54) (1.38) (0.85) (1.22) (1.41)
Ashamed 1.43 1.29 1.17 1.29 1.41 1.79 1.77 1.36
(0.53) (0.49) (0.38) (0.81) (0.91) (1.35) (1.31) (0.92)
2.14 1.86 2.22 1.64 1.69 2.42 2.05 1.45
Nervous
(1.21) (0.69) (1.63) (1.13) (1.00) (1.64) (1.50) (1.21)
Jitte 1.57 1.57 2.33 2.14 2.53 2.46 2.77 1.45
ry (0.79) (1.13) (1.68) (1.65) (1.80) (1.64) (1.74) (0.93)
Afraid 1.29 1.43 1.44 1.21 1.28 1.50 1.36 1.00
(0.49) (0.79) (0.86) (0.63) (0.52) (1.14) (0.58) (0.00)

 

78

 

 

Table 7

T -tests of Categorical Coded Variables Among Confronters

 

Coded Variable Dependent Variable

 

Participant Positive Affect Participant Negative Affect

 

 

 

 

 

 

1(dfi M (SD) t (df) M (SD)
Evaluation of Comment
Neutral 3.64 (1.27) 1.16 (0.59)
Negative 3.50 (1.16) 2.14 (0.89)
0.44 (58) -2.77* (58)
Expressed Uncertainty
Did not express 3.31 (1.26) 1.89 (0.81)
Did express 4.07 (1.02) 1.68 (0.68)
-2.32* (59) 0.97 (59)
Made a Suggestion
Did not suggest 3.54 (1.39) 1.76 (0.75)
Did suggest 3.56 (1.01) 1.95 (0.79)
-0.80 (59) -l.l3 (59)
Labeled Comment
Did not label 3.50 (1.25) 1.61 (0.62)
Did label 3.64 (1.22) 2.20 (0.88)
-0.44 (59) -3.07* (59)
Labeled Person
Did not label 3.55 (1.20) 1.81 (0.77)
Did label 3.54 (1.55) 1.90 (0.81)
0.01 (59) -0.28 (59)

 

Tp<.10,*p<.05

79

Table 8

Correlations and Descriptive Statistics of Continuous Coded Variables and Participant
Affect Variables Among Confronters

 

 

 

Reaction to . . . Essay
comment Frrendly Helpful Crrtrcal length
Participant *
Positive Affect -0.03 0.29 0.00 -0.19 0.20
Participant *
Negative Affect 0.28 -0.25 0.14 0.23 0.15
Means 2.09 2.39 2.23 2.66 150.91
S’a’fafd 1.29 0.89 1.03 1.01 61.77
Devratrons

 

Tp<.10,*p<.05

80

Table 9

Means and Standard Deviations of Partner Aflect Variables by Condition, Sex, and
Confronter Status

 

 

 

 

Confronted Did not confront
Rude Sexist Rude Sexist
M W M W M W M W
PA 2.79 2.69 3.26 2.88 3.37 3.15 3.15 2.49
(0.81) (0.88) (0.74) (1.30) (1.15) (1 .01) (1.24) (0.76)
NA 2.39 2.36 2.67 2.72 2.22 2.43 2.07 2.14

(1.10) (1.02) (1.41) (1.19) (0.99) (1.14) (0.82) (1.00)

 

81

Table 10

Means and Standard Deviations of Partner Affect Individual Items by Condition, Sex, and
Confronter Status

 

 

 

 

Confronted Did not confront
Rude Sexist Rude Sexist
M W M W M W M W
Partner Positive Affect Items

Interested 3.29 3.71 4.06 4.18 3.81 3.67 3.50 3.45
(2.06) (0.95) (1.30) (1 .3 3) (1.28) (1.46) (1.47) (0.93)

Excited 2.71 2.29 2.83 2.57 2.90 2.88 2.68 2.09
(1.1 1) (0.76) (1 .25) (1 .50) (1.35) (0.90) (1 .46) (0.83)

Stron 3.43 2.29 3.33 2.43 3.16 3.00 3.00 2.18
g (1.72) (1.25) (1.28) (1.53) (1.48) (1.32) (1.41) (0.98)

Enthusiastic 2.14 1.86 3.33 2.36 3.03 3.04 2.95 2.27
(1 .21) (0.90) (1 .68) (1.64) (1.53) (1.37) (1.53) (1.10)

Proud 2.29 2.86 3.28 2.21 3.44 2.83 3.23 1.82
(1 .25) (1.35) (1.07) (1.52) (1 .44) (1.43) (1.69) (0.98)

Alert 3.43 2.86 3.56 3.46 3.19 3.50 3.00 3.00
(1 .27) (1 .46) (1.34) (1.99) (1 .60) (1.47) (1.60) (1.61)

Ins ired 2.00 2.71 2.61 2.54 3.63 3.04 2.82 2.45
p (1.00) (0.76) (1.24) (1.64) (1.43) (1 .33) (1.62) (1.43)

Determined 2.57 3.14 2.56 2.89 3.09 2.96 3.09 2.18
(1.51) (1.57) (1.34) (1.87) (1.53) (1.30) (1.41) (1.32)

Attentive 3.57 3.00 3.78 3.57 4.00 3.63 3.95 3.27
(1.72) (1.41) (1.40) (1.91) (1.31) (1.41) (1.50) (1.62)

Active 2.43 2.14 3.28 2.57 3.41 2.96 3.23 2.18
(1 .40) (1 .68) (1.36) (1.71) (1.58) (1.46) (1.54) (1.25)

 

82

Table 10 (cont’d)

 

 

 

 

Confronted Did not confront
Rude Sexist Rude Sexist
M W M W M W M W
Partner Negative Affect Items

Distressed 2.86 2.71 2.89 3.00 2.28 2.54 2.50 2.18
(1.46) (1.11) (1.60) (1.63) (1.25) (1.53) (1.50) (1.25)

U set 3.00 2.86 3.22 3.36 2.75 2.75 2.32 2.91
p (1.73) (1.35) (1.70) (1.25) (1.37) (1 .73) (1.09) (1.22)

Guilt 1.29 2.57 2.56 3.36 1.97 2.17 1.81 1.81
y (0.76) (1.90) (1.76) (1.64) (1 .47) (1.49) (1.01) (1.47)

An 3.00 2.29 2.83 3.04 2.69 2.67 2.23 2.72

gry (2.00) (1.38) (1 .72) (1 .69) (1.47) (1.71) (1.15) (1.74

Hostile 3.00 2.29 2.61 2.79 2.19 2.54 1.77 1.82
(1.91) (1.60) (1.65) (1.75) (1.33) (1.53) (0.97) (1.66)

Irritable 2.71 2.57 2.67 3.04 2.50 2.63 2.27 2.73
(1 .70) (0.79) (1.88) (1 .67) (1.22) (1.38) (1.24) (1.50)

Ashamed 1.86 2.43 2.83 2.86 2.09 2.13 1.91 2.09
(0.90) (1.13) (1.72) (1.90) (1.25) (1.39) (0.92) (1.58)

Nervous 2.57 2.29 2.67 2.07 2.06 2.96 2.00 1.82
(1.27) (1.50) (1.61) (1.56) (1 .24) (1.78) (0.82) (0.98)

Jitte 2.00 2.14 2.44 1.93 2.19 2.25 2.32 1.91
ry (1.53) (1.35) (1.50) (1.33) (1.23) (1.45) (1 .39) (1.04)

Afraid 1.57 1.43 1.94 1.75 1.47 1.71 1.59 1.36
(0.96) (0.79) (1.26) (1 .29) (0.88) (1.04) (1.01) (0.50)

 

83

Table 1 1

T -tests of Categorical Coded Variables Among Confronters

 

Coded Variable Dependent Variable

 

Participant Positive Affect Participant Negative Affect

 

 

 

 

 

 

t (df) M (SD) t (d)? M (SD)
Evaluation of Comment
Neutral 3.09 (1.07) 2.26 (0.96)
Negative 2.84 ( 1.04) 3.22 (1.33)
0.92 (58) -3.21* (58)
Expressed Uncertainty
Did not express 2.94 (1.12) 2.74 (1.19)
Did express 3.04 (0.90) 2.45 (1.31)
-0.03 (59) 0.86 (59)
Made a Suggestion
Did not suggest 2.93 (1.18) 2.32 (1.22)
Did suggest 3.03 (0.86) 3.10 ( 1.10)
-0.40 (59) -2.55* (59)
Labeled Comment
Did not label 2.92 (1.05) 2.42 (1.05)
Did label 3.06 (1.07) 3.07 (1.24)
-0.51 (59) -2.07* (59)
Labeled Person
Did not label 2.94 (1.06) 2.54 (1.20)
Did label 3.26 (1.00) 3.56 (1.08)
-0.76 (59) -2.14* (59)

 

Tp<.10,*p<.05

84

Table 12

Correlations and Descriptive Statistics of Continuous Coded Variables and Partner

 

 

 

Aflect
Reaction Essa
to Friendly Helpful Critical y
length
comment
Partner * * *
Positive Affect 0.09 0.30 0.03 —0.26 0.28
Partner * * * *
Negative Affect 0.44 -0.38 0.21 0.46 0.39
Means 2.09 2.39 2.23 2.66 150.91
Standard Deviations 1.29 0.89 1.03 1.01 61.77

 

Tp<.10,*p<.05

85

Table 13

Means and Standard Deviations of Partner Evaluation Variables by Condition, Sex, and
Confronter Status

 

 

 

 

Confronted Did not confront
Rude Sexist Rude Sexist
M W M W M W M W
Positive 3.16 3.47 3.71 3.64 3.71 3.61 3.48 3.06
Evaluation (0.86) (1.02) (0.77) (1.21) (0.95) (0.70) (1 .01) (1.06)
Negative 2.87 2.49 2.92 2.89 2.85 2.67 2.69 2.62

Evaluation (1.52) (1.17) (1.43) (1.15) (1.28) (1.18) (1.20) (1.07)

 

86

Table 14

Means and Standard Deviations of Participant A fleet Variables by Condition, Sex, and

 

 

 

 

Confronter Status
Confronted Did not confront
Rude Sexist Rude Sexist
M W M W M W M W
Partner Positive Evaluation Items
Justified 3.14 3.29 3.78 3.57 3.91 4.00 3.36 3.45
(0.90) (0.95) (1.35) (1.45) (1.17) (0.98) (1.40) (1.13)
Making a 3.57 3.86 4.33 4.32 4.16 4.29 3.95 3.73
good point (0.98) (1 .07) (0.77) (1.56) (1.48) (1.08) (1.29) (1.68)
Interestin 2.71 3.14 3.71 3.32 3.78 3.38 3.45 3.09
g (1.25) (0.90) (1.60) (1.59) (1.34) (1.31) (1.14) (1.58)
3:353; 3.71 3.29 3.72 3.93 3.19 2.96 3.00 2.45
issfies (1.38) (2.14) (1.41) (1.80) (1.18) (1.12) (1.31) (1.75)
Enlightened 2.29 2.57 2.94 2.61 2.91 2.71 3.00 1.01
(1 .38) (1 .27) (1.66) (1.59) (1 .30) (1.20) (1 .45) (0.70)
Being 3.57 4.29 3.67 4.00 4.19 4.08 3.91 3.45
reasonable (1 .27) (1 .60) (1.28) (1 .70) (1.42) (1 .06) (1 .57) (1 .21)
He] fill 3.14 3.86 3.83 3.71 3.84 3.88 3.68 3.36
p (1.35) (1.46) (1.10) (1.90) (1.42) (1.19) (1.52) (1.02)

 

87

Table 1.4 (cont’d)

 

 

 

 

Confronted Did not confront
Rude Sexist Rude Sexist
M W M W M W M W
Partner Negative Evaluation Items
Being hyper- 2.86 2.71 3.11 3.50 3.06 3.13 2.54 2.45
sensitive (1.68) (1.80) (1.86) (1.20) (1.88) (1.94) (1.41) (1.37)
Overreactin 3.57 3.00 3.11 3.50 3.28 3.04 2.45 3.54
g (2.22) (1.83) (1.94) (1.750 (1 .90) (1.85) (1.50) (1.63)
Self- 2.57 2.43 3.22 3.61 3.41 2.96 3.32 2.82
righteous (1.51) (1 .27) (1 .80) (1.71) (1.72) (1.37) (1.32) (1 .72)
Im olite 2.86 2.29 2.89 2.46 2.72 2.71 2.55 2.73
p (1.86) (1.11) (1.49) (1.75) (1.65) (1.73) (1.44) (1.35)
36mg 3.00 1.86 2.50 2.75 2.66 2.58 2.23 2.09
unreason-
able (1.91) (0.69) (1.58) (1 .38) (1.62) (1.56) (1.45) (1.04)
A 3.00 2.71 2.78 2.61 2.78 2.67 2.50 2.73
complainer (2.00) (1 .60) (1.86) (1.34) (1.83) (1.66) (1.65) (1.27)
Anno in 3.14 3.00 3.17 2.64 2.66 2.92 2.27 2.73
y g (2.41) (1.41) (1.76) (1.54) (1.62) (1.32) (1.35) (1.56)
Emotional 2.71 2.43 2.83 2.86 2.75 2.08 2.14 1.91
(1.60) (1.13) (1 .86) (1.51) (1.61) (1.10) (0.99) (0.94)
Acting too 2.14 2.00 2.78 2.93 2.31 1.96 2.05 1.45
liberal (1.68) (1.15) (1 .80) (1 .54) (1.33) (1.33) (1.17) (0.69)

 

88

Table 15

T-tests of Categorical Coded Variables Among Confronters

 

Coded Variable Dependent Variable

 

Participant Positive Affect Participant Negative Affect

 

 

 

 

 

 

t (df) M (SD) t (df) M (SD)
Evaluation of Comment
Neutral 3.31 (1 .07) 2.74 (1.42)
Negative 3.97 (0.84) 3.28 (1.10)
-2.57* (58) -l .59 (58)
Expressed Uncertainty
Did not express 3.56 (1.07) 3.02 (1 .27)
Did express 3.62 (0.92) 2.78 (1 .40)
-0.20 (59) 0.67 (59)
Made a Suggestion
Did not suggest 3.47 (1 .03) 2.74 (1 .40)
Did suggest 3.73 (1.00) 3.23 (1.15)
-0.98 (59) -1.46 (59)
Labeled Comment
Did not label 3.28 (1 .04) 2.91 (1 .49)
Did label 4.10 (0.73) 3.00 (0.92)
-3.26* (59) -0.24 (59)
Labeled Person
Did not label 3.56 (0.96) 2.82 (1 .29)
Did label 3.73 (1 .49) 3.94 (1.07)
-0.43 (59) -2.20* (59)

 

p<.10,*p<.05

89

Table 16

Correlations and Descriptive Statistics of Continuous Coded Variables and Partner
Evaluation Variables Among Confronters

 

 

Reaction Essa
to Friendly Helpful Critical y
length
comment
Partner .Posmve 0.23 038* 0.08 -0. 15 029*
Evaluation
Partner .Negatwe 032* -0.32* 0.24 049* 032*
Evaluation
Means 2.09 2.39 2.23 2.66 150.91
Stanfield 1.29 0.89 1.03 1.01 61 .77
Devratlons

 

Tp<.10,*p<.05

90

 

Table 17

Means and Standard Deviations of Effectiveness Variable by Condition, Sex, and
Confronter Status

 

Confronted Did not confront

 

Rude Sexist Rude Sexist

 

M W M W M W M W

 

2.89 3.46 3.47 2.74 2.64 3.04 2.55 3.16

EffeCtive (0.62) (0.57) (1.01) (1.51) (1.03) (0.95) (0.86) (1.24)

 

91

 

 

Table 18

T -tests of Categorical Coded Variables Among Confronters

 

Coded Variable Dependent Variable

 

Participant Positive Affect

 

t (df) M (SD)

 

Evaluation of Comment

 

 

 

 

Neutral 3.14 (1.10)
Negative 4.09 (1 .20)
-3.19* (58)
Expressed Uncertainty
Did not express 3.65 (1.22)
Did express 3.29 (1.19)
1.05 (59)
Made a Suggestion
Did not suggest 3.31 (1.11)
Did suggest 3.85 (1.29)
-1 .781 (59)
Labeled Comment
Did not label 3.38 (1.16)
Did label 3.81 (1.29)
-1.32 (59)
Labeled Person
Did not label 3.47 (1 .08)
Did label 4.09 (2.01)
-1.28 (59)

 

'l'p<.10,*p<.05

92

 

Table 19

Correlations and Descriptive Statistics of Continuous Coded Variables and Perceived
Essay Effectiveness Among Confronters

 

 

 

Reaction Essa
to Friendly Helpful Critical y
length
comment
Essay Effectiveness 0.18 -0.18 0.18 0.16 0.17
Means 2.09 2.39 2.23 2.66 150.91
Standard Deviations 1.29 0.89 1.03 1.01 61 .77

 

'l‘p<.10,*p<.05

93

References

Allport, G. W. (1954). The nature of prejudice. Reading, MA: Addison-Wesley.

Batson, D. C., Polycarpou, M. P., Harmon-Jones, E., Imhoff, H. J ., Mitchener, E. C.,
Bednar, L. L., Klein, T. R., & Highberger, L. (1997). Empathy and attitudes: Can
feeling for a member of a stigmatized group improve feelings toward the group?
Journal of Personality and Social Psychology, 72, 105-1 18.

Channouf, A., Py, J ., & Somat, A. (1999). Cognitive processing of causal explanations:
A sociocognitive perspective. European Journal of Social Psychology, 29, 673-
690.

Czopp, A. M., & Monteith, M. J. (2003). Confronting prejudice (literally): Reactions to
confrontations of racial and gender bias. Personality and Social Psychology
Bulletin, 29, 532-544.

Czopp, A. M., & Monteith, M. J. (2006). Standing up for change: Reducing bias through
interpersonal confrontation. Journal of Personality and Social Psychology, 90,
784-803.

Devine, P. G., Monteith, M. J ., Zuwerink, J. R., & Elliot, A. J. (1991). Prejudice with and
without compunction. Journal of Personality and Social Psychology, 60, 817-830.

Donnellan, M. B., Oswald, F. L., Baird, B. M., Lucas, R. E. (2006). The Mini-IPIP
Scales: Tiny-yet-effective measures of the Big Five factors of personality.
Psychological Assessment, 18, 192-203.

Dovidio, J. F., Gaertner, S. L., Validzic, A. (1998). I ntergroup bias: status,
differentiation, and a common in-group identity. Journal of Personality and
Social Psychology, 75, 109-120.

F estinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117-
140.

Garcia, D. M., Reser, A. H., Arno, R. B., Redersdorff, S., & Branscombe, N. R. (2005).
Perceivers' Responses to In-Group and Out-Group Members Who Blame a

Negative Outcome on Discrimination. Personality and Social Psychology
Bulletin, 31, 769-780.

Gigerenzer, G., & Todd, P. M. (1999). Fast and frugal heuristics: The adaptive toolbox.

The ABC Research Group Simple heuristics that make us smart. New York:
Oxford University Press.

94

Glick, P., & Fiske, S. T. (1996). The Ambivalent Sexism Inventory: Differentiating
hostile and benevolent sexism. Journal of Personality and Social Psychology, 70,
491 -5 12.

Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect. Psychological
Review, 94, 319-340.

Homsey, M. J., & Hogg, M. A. (2000). Subgroup relations: A comparison of mutual
intergroup differentiation and common ingroup identity models of prejudice
reduction. Personality and Social Psychology Bulletin, 26, 242-256.

Kaiser, C. R., & Miller, C. T. (2001). Stop complaining! The social costs of making
attributions to discrimination. Personality and Social Psychology Bulletin, 27,
254-263.

Kaiser, C. R., & Miller, C. T. (2003). Derogating the victim: The interpersonal
consequences of blaming events on discrimination. Group Processes and
Intergroup Relations, 6, 227-237.

Kaiser, C. R., & Miller, C. T. (2004). A Stress and Coping Perspective on Confronting
Abstract Sexism. Psychology of Women Quarterly, 28, 168-178.

Klonis, S. C., Plant, E. A., & Devine, P. G. (2005). Internal and External Motivation to
Respond Without Sexism. Personality and Social Psychology Bulletin, 31, 123 7-
1249.

Lazarus, R. S. (1999). Stress and Emotion: A new synthesis. New York: Springer.

Leary, M. R., & Kowalski, R. M. (1990). Impression management: A literature review
and two-component model. Psychological Bulletin, 107, 34-47.

Marcus-Newhall, A., Miller, N., Holtz, R., & Brewer, M. B. (1993). Cross-cutting
category membership with role assignment: A means of reducing intergroup bias.
British Journal of Social Psychology, 32, 125-146.

Monteith, M. J. (1993). Self-regulation of prejudiced responses: Implications for progress
in prejudice-reduction efforts. Journal of Personality and Social Psychology, 65,
469-485.

Operario, D., & F iske, ST (2001). Ethnic identity moderates perceptions of prejudice:
Judgments of personal versus group discrimination and subtle versus blatant bias.
Personality and Social Psychology Bulletin, 27, 550-561.

Petty, R. E., Fleming, M. A., Priester, J. R., & Feinstein, A. H. (2001). Individual versus

group interest violation: Surprise as a determinant of argument scrutiny and
persuasion. Social Cognition, 19, 418-442.

95

 

Sellers, R. M., Rowley, S. A. J ., Chavous, T. M., Shelton, J. N., & Smith, M. A. (1997).
Multidimensional Inventory of Black Identity: A preliminary investigation of

reliability and construct validity. Journal of Personality and Social Psychology.
73, 1-1 1.

Sellers, R. M., & Shelton, J. N. (2003). The role of racial identity in perceived racial
discrimination. Journal of Personality and Social Psychology, 84, 1079-1092.

Shelton, J. N., & Richeson, J. A. (2005). Intergroup Contact and Pluralistic Ignorance.
Journal of Personality and Social Psychology, 88, 91-107.

Shelton, J. N., Richeson, J. A., Salvatore, J ., & Hill, D. M. (2005). Silence is not always
golden: Intrapersonal consequences of not challenging prejudice. In S. Levin and
C. Van Laar (Eds), Social stigmas. Erlbaum.

Shelton, J. N., & Stewart, R. E. (2004). Confronting perpetrators of prejudice: The
inhibitory effects of social costs. Psychology of Women Quarterly, 28, 215-223.

Smith, H. J ., & Tyler, T. R. (1996). Justice and Power: When will justice concerns
encourage the advantaged to support policies which redistribute economic

resources and the disadvantaged to willingly obey the law? European Journal of
Social Psychology, 26, 171-200.

Snyder, M., & Gangestad, S. (1986). On the nature of self-monitoring: Matters of

assessment, matters of validity. Journal of Personality and Social Psychology, 51,
125-139. ~

Stephan, W. G. & Stephan, C. (1985). Intergroup anxiety. Journal of Social Issues, 41,
157-176.

Swim, J. K., Aikin, K. J ., Hall, W. S., & Hunter, B. A. (1995). Sexism and racism: Old-
fashioned and modern prejudices. Journal of Personality and Social Psychology,
68, 199-214.

Swim, J. K., & Hyers, L. L. (1999). Excuse me— What did you just say?!: Women's public

and private responses to sexist remarks. Journal of Experimental Social
Psychology, 35, 68-88.

Tropp, L. R., & Pettigrew, T. F. (2005). Differential Relationships Between Intergroup
Contact and Affective and Cognitive Dimensions of Prejudice. Personality and
Social Psychology Bulletin, 31. 1 145-1 158.

Vescio, T. K., Sechrist, G. B., & Paolucci, M. P. (2003). Perspective taking and prejudice

reduction: The mediational role of empathy arousal and situational attributions.
European Journal of Social Psychology, 33, 455-472.

96

 

 

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief
measures of positive and negative affect: The PANAS scales. Journal of
Personality and Social Psychology, 54, 1063-1070.

Winslow, M. P. (2004). Reactions to the Imputation of Prejudice. Basic and Applied
Social Psychology, 26, 289-297.

Woodzicka, J. A., & LaFrance, M. (2001). Real versus imagined gender harassment.
Journal of Social Issues, 5 7, 15-30.

Wright, S. C., Aron, A., McLaughlin Volpe, T., & Ropp, S. A. (1997). The extended

contact effect: Knowledge of cross-group friendships and prejudice. Journal of
Personality and Social Psychology, 73, 73-90.

97

   

11111111111111111111111911151111