Social Identity Lab Research
In my research, I seek to understand the interplay between self and social identity, particularly when one's social identity is accorded lower status or is targeted by negative stereotypes. In exploring these issues, my research draws upon and extends existing research on social stigma, social justice, social cognition, intergroup emotion, self-esteem, stress and arousal, attention and working memory, and motivation and performance. My research examines three related questions:
1) What variables moderate, mediate, and alleviate the inhibiting effects of stereotype threat on an individual's performance? How does stereotype threat affect learning?
2) How and when do members of lower status groups psychologically disengage their self-esteem from outcomes they receive in stereotype relevant domains?
3) When and why do people feel ashamed and/or guilty for the negative actions of their ingroup?
Stereotype Threat - What Moderates it, Mediates it, and Alleviates it?
One primary line of research in our lab involves an investigation of variables
that play a role in the inhibiting effects of negative group stereotypes
on academic test performance. Steele and his colleagues (1997; Steele
& Aronson, 1995) have demonstrated that members of socially stigmatized
groups perform more poorly than nonstigmatized others when negative stereotypes
about their ingroup are made salient. One goal of our research has been
to identify the variables that mediate and moderate these effects. For
example, in two separate papers, we report evidence that a manipulation
of stereotype threat only inhibits women's math performance if gender
is an important aspect of their self-definition (Schmader, 2001) and if
women have at least a suspicion that the stereotype might be true (Schmader,
Johns, & Barquissau, 2004). Thus, just as group identification and
the perceived legitimacy of lower status influences the types of self-protective
strategies used by members of stigmatized groups, it also appears to influence
the effect that group stereotypes have on their performance.
Our research has also tested variables that mediate stereotype threat
effects. I am particularly interested in the combined roles of affective,
cognitive, and physiological mediators. A central focus of this research
has been to show that manipulations of stereotype threat inhibit performance
by reducing working memory capacity - the ability to hold information
in memory while focusing attention on some other cognitive task. In a
series of three experiments (Schmader & Johns, 2003), my graduate
student and I have shown that women and Latinos experience reduced working
memory capacity under conditions of stereotype threat that inhibits their
ability to perform well on complex intellectual tasks. This finding suggests
that when negative self-relevant stereotypes are made salient, individuals'
attentional resources are being divided between the task in front of them
(e.g., taking a test) and some other cognitive process.
We are currently testing a model of the basic psychological processes
that might underlie the effects of stereotype threat on working memory
capacity (Schmader, Johns, & Forbes, 2008). These include processes like performance monitoring and anxiety
suppression. In a recent set of studies (Johns, Inzlicht, & Schmader, 2008),
we have found evidence that women experience reduced working memory capacity
under stereotype threat to the degree that they try to suppress or avoid
feelings of anxiety they might be experiencing. This research assumes that anxiety relates to reduced capacity and poorer performance because people interpret their anxiety as a sign of failure. To test this idea, other current research shows that anxiety is particularly related to lower working memory and performance when it is experienced in the context of negative thoughts (Schmader, Forbes, Zhang, & Mendes, 2008). When anxiety is paired with positive thoughts or is reappraised as being a more positive experience, greater arousal in a situation of stereotype threat actually predicts better performance.
In addition to examining variables that moderate and mediate stereotype threat effects on performance,
we have also begun research addressing how stereotype threat can be alleviated.
In one study, we have evidence that teaching women that stereotype threat
is an explanation for why they might experience increased anxiety or arousal
during a math test actually inoculates them against its effects (Johns,
Schmader, & Martens, 2005). Taken together, our research is showing that changing how people interpret their psychological experience in situations of stereotype threat can be an important strategy for inoculating people against its effects.
In future research, we will be moving away from examining how stereotype threat affects performance, to explore how situations of threat impact the process of learning new information. Preliminary evidence suggests that situations of threat might bias people toward negative feedback and that under certain circumstances, learning can actually be facilitated by stereotype threat, even though individuals might be impaired in their ability to demonstrate that learning.
The Self-Protective Processes of Psychological Disengagement
My second program of research, explores the processes whereby individuals
detach their sense of self-worth from the outcomes they receive in a given
domain, a process we refer to as psychological disengagement. Our theoretical
model of psychological disengagement includes two separate pathways by
which disengagement may occur (Schmader, Major, & Gramzow, 2001; Major
& Schmader, 1998). The first is a process of devaluing the importance
of success in a domain. The second is a process of discounting outcomes
or evaluations as being inaccurate representations of one's abilities.
Although anyone might disengage his or her self-esteem to protect against
negative outcomes, my research concerns the occurrence of these processes
among members of lower status groups, and in particular, ethnic minorities
who might disengage their self-esteem from academic outcomes (Schmader
et al., 2001; Major & Schmader, 1998; Major, Spencer, Schmader, Wolfe,
& Crocker, 1998). Our research reveals that Black and Latino/a college
students are more likely than White students to have psychologically disengaged
their self-esteem from their academic outcomes (Schmader, Major, &
Gramzow, 2001; Major & Schmader, 1998; Major, Spencer, Schmader, Wolfe,
& Crocker, 1998). Additional findings suggest that Black students
may be more disengaged from academics because they discount their test
scores as biased and not because they devalue the importance of academic
success.
We have conducted additional research that has focused more specifically
on individual difference variables that moderate both devaluing and discounting.
One line of studies examines the role of perceived legitimacy in moderating
devaluing. In an earlier paper, I had shown that in the absence of personal
experience in a novel domain, individuals devalue a domain to the extent
that their ingroup is unsuccessful in that domain relative to an outgroup
(Schmader & Major, 1999). However, an additional set of studies shows
this tendency to devalue group failure is moderated both by the relative
status of one's ingroup and the perceived legitimacy of that status (Schmader,
Major, Eccleston, & McCoy, 2001). When the status hierarchy is seen
as legitimate, members of higher status groups devalue domains in which
lower status groups excel, whereas members of lower status groups actually
value domains in which higher status groups excel. However, members of
lower status group do devalue domains in which higher status groups outperform
their own group if the status hierarchy defining the groups is delegitimized.
We have found these effects with groups based on school affiliation, gender,
and ethnicity. We have also discussed the role of perceived legitimacy
in how people cope with social stigma in Schmader et al. (2001) and Major
and Schmader (2001).
In other studies, we have investigated individual difference variables
that moderate discounting processes more specifically (Schmader, Major,
& Gramzow, 2001; Major, Gramzow, McCoy, Levin, Schmader, & Sidanius,
2002). This research shows that members of lower status groups are less
likely to discount a negative outcome as discrimination and are less likely
to discount academic tests as being biased against them to the extent
that they endorse ideologies that justify existing group differences.
Additional research suggests that individual differences in gender identification
also moderate discounting processes and the extent to which individuals
interpret ambiguous situations of bias as discrimination (Major, Quinton,
& Schmader, 2002).
In our most recent research, we are examining neurocognitive correlates of disengagement due to stereotype threatening situations. Our results suggest that minority students who devalue academics show a lower ERN amplitude (a component of their ERP response) to mistakes that they make during a task that they believe indexes intelligence, whereas minority students who value academics show a stronger ERN amplitude to mistakes (Forbes, Schmader, & Allen, 2008). In contrast, a tendency to discount academic feedback does not predict initial attention to errors, but instead moderates how those errors are evaluated and the subjective feelings of threat in the situation. These data begin to point to the importance of performance monitoring and evaluation to the learning process and the role that this plays in minority students' achievement. We are continuing to examine neurocognitive correlates of attention and memory encoding under stereotype threat and how they related to subsequent learning and performance.
Vicarious Shame and Guilt felt for the Actions of Ingroup Members
In my third line of research, conducted in collaboration with Brian Lickel
at USC, I am studying when and why people feel ashamed or guilty for negative
actions of their ingroup. Earlier work suggests that although shame and
guilt are related emotions, they are nonetheless psychologically distinct,
and are characterized by differing antecedent appraisals and motivational
consequences (e.g., Tangney, 1995; Wicker, Payne, & Morgan, 1983).
Specifically, people feel guilty when an event is seen as implicating
something negative about their behavior, whereas people feel ashamed when
an event is seen as implicating something negative about their sense of
self. Interestingly, very little work has investigated instances in which
these emotions are prompted not by one's own wrongdoings, but by the actions
of others. It is critical to distinguish between vicarious guilt and vicarious
shame because whereas guilt tends to elicit a desire to repair whatever
harm has been done, shame tends to elicit a desire to hide or distance
oneself from the negative event.
We have developed a theoretical model of vicarious shame and guilt that
integrates conceptual distinctions between self-conscious emotions and
between social associations to make predictions about the types of appraisals
that underlie vicarious shame and guilt (Lickel, Schmader, Curtis, Barquissau,
& Ames, 2004; Lickel, Schmader, & Barquissau, 2007). In 2001,
we received a three-year grant from the National Science Foundation to
fund our work in this area. In one set of studies (Lickel et al., 2004),
we have found that people feel guilty for another's wrongdoing to the
extent that they have a highly interdependent association to that wrongdoer
(e.g., one that involves a great deal of interpersonal interaction) and
thus feel that they should have been able to control that person's actions.
In contrast, people feel ashamed for another's wrongdoing to the extent
that the person's behavior is seen as relevant to sense of social identity
shared in common with the wrongdoer and thus appraise the other person's
behavior as a reflection of their own self-image.
Our model can be applied to any interpersonal association. For example, parents who report feeling ashamed for the wrongdoings of their children are more likely to respond to those events by distancing themselves from their child, withholding warmth from them, and using overreactive forms of punishment (Scarnier, Schmader, & Lickel, 2008). In contrast, feeling guilty for a wrong committed by a child predicts wanting to apologize for what the child has done and use more inductive forms of parenting that help the child understand why his or her behavior is wrong.
In addition to applying this model to interpersonal relationships, ,
I am particularly interested in people's emotional reactions in intergroup
situations. We have evidence from two samples of Latino/a students that
when they see other members of their ethnic group doing something wrong,
they feel higher levels of shame (but not guilt) to the extent that they
see those behaviors as confirming social stereotypes about their group.
These feelings of shame then predict a desire to distance themselves not
only from the stereotypic group member, but also from their ethnic group.
Feelings of vicarious shame also predict engaging in behaviors that will
disconfirm the stereotype. These studies and our general ideas about stigma
and shame are summarized in a recent book chapter (Schmader & Lickel,
2006).
Expanding upon experiences of vicarious shame and guilt felt by those who are socially stigmatized, my graduate students and I have explored vicarious shame and
guilt experienced by White students who witness another member of their
race confirm the stereotype that Whites are racist. Our model predicts
that nonprejudiced White students should experience "White shame,"
not "White guilt," and this affective reaction should evoke
withdrawing behaviors rather than approach efforts. For example, one set
of studies reveal that the emotional reaction (either anger or sadness)
of those who are victimized by an ingroup member's prejudice moderate
the intensity of White shame (Barquissau, Schmader, & Lickel, in preparation).
In a related study (Johns, Schmader, & Lickel, 2004), we have found
that Americans report feeling ashamed, and not guilty, for the extreme
acts of prejudice committed by other Americans against Arab- and Muslim-appearing
individuals in the wake of the 9-11 terrorist attacks. These feelings
of shame are more intense among those who are highly identified with being
American and predict a desire to distance oneself from the events and
even from their national identity.
Copyright ® 2004 BurningPixels Designs