As a social psychologist in the tradition of Kurt Lewin, I am committed to identifying psychological processes that contribute to social problems and to developing theory-based interventions to affect these processes. My research examines diverse contexts, including education, health, intergroup relations, politics, and the environment using both laboratory and field-experimentation. This research simultaneously advances psychological theory, demonstrates the importance of psychological processes in major social problems, and suggests novel remedies to these problems.

A Social-Psychological Approach to Group Differences in Academic Achievement

One of the most pressing problems society faces involves the persistence of group differences in academic achievement between ethnic and gender groups. While many structural factors contribute to inequality in education, I have long been interested in the role of psychological factors, for instance factors that originate from negative intellectual stereotypes. In early research, I investigated stereotype lift, the performance boost experienced by members of non-stereotyped groups like Whites and men when they know that an outgroup (e.g., racial minorities, women) is negatively stereotyped in an academic setting (Walton & Cohen, 2003). More recently, I have investigated how stereotypes and numeric underrepresentation negatively affect the experience, motivation, and achievement of minority-group students in school settings, and how theory-based interventions that address these processes can reduce inequality in education (for a review, see Yeager & Walton, 2011).

Social Belonging and Group Differences in Academic Achievement

A fundamental theoretical emphasis of my research involves the importance of a sense of social belonging in human functioning in general and in motivation and achievement in particular. Yet one consequence of negative stereotypes is to cause students to wonder whether others in school will fully include and value them. In this state of belonging uncertainty (Walton & Cohen, 2007), students may monitor the school environment for indicators of whether they belong or do not. From this perspective, even subtle events—like critical feedback or feelings of loneliness—may seem diagnostic of a global lack of belonging in the school. In turn, inferring a lack of belonging may undermine students’ motivation and academic performance. Collaborating with Geoff Cohen, I found evidence for this process in laboratory research. We exposed Black and White students to subtle information indicating that they might have few friends in a field of study. While the manipulation had no effect on White students, it led Black students to question their belonging and their potential in the field, and even to discourage a same-race peer from pursuing the field (Walton & Cohen, 2007).

To prevent such deleterious attributions, we then created a social belonging intervention (Walton & Cohen, 2011). This intervention provides students a nonthreatening explanation for feelings of nonbelonging in general in school. The intervention was first tested in a sample of Black and White first-year college students. Students read a survey of upper-year students at their school. The survey indicated that feelings of nonbelonging are normal at first in college for students of all ethnicities and dissipate with time. The information was designed to lead students to attribute negative events and feelings of nonbelonging to the difficulty of the transition to college rather than to a lack of belonging on their part or on the part of their racial group (cf. Wilson et al., 2002). An additional writing exercise encouraged students to internalize the treatment message. In total, the intervention lasted about an hour.

For White students, who have little cause to doubt their belonging in school on account of their race, the intervention had little impact. But for Black students, the intervention delivered in the spring of their freshman year raised GPA from sophomore-through-senior year as compared to multiple control groups. Overall, from sophomore-through-senior year, the intervention cut the achievement gap between Black and White students by half (Walton & Cohen, 2011). How did this occur? Evidence suggests that the predicted effect on students’ social construal mediated the long-term gains in performance. Daily diaries completed soon after the intervention showed that the intervention sustained Black students’ sense of belonging in the face of adversity. No longer did negative events seem to carry a symbolic, threatening meaning. These findings simultaneously address an important social problem and advance psychological theory.

Social belonging is an especially important target for intervention in part is because it is a fundamental need that contributes broadly to human functioning (see Baumeister & Leary, 1995). Indeed, social belonging may serve as a “psychological hub” with links to diverse outcomes. If so, interventions that buttress belonging may have broad effects. To examine this, at the end of students’ college tenure (3-years postintervention) a survey assessed well-being and physical health. Results showed that the intervention eliminated race differences in happiness and in self-assessed health (Walton & Cohen, 2011). Whereas 60% of untreated Black students had seen a doctor recently, only 28% of treated Black students had. These findings suggest the potential for psychological interventions to address racial disparities in health, which is an alarming problem even among people with the same socioeconomic background. To determine whether these patterns generalize to objective indicators of health, ongoing research with Dr. Steve Cole examines effects on students’ health records, for instance in terms of the incidence of infectious disease. To further understand the role of identity threats in racial disparities in health and the effectiveness of psychological remedies, we hope to include physiological measures in future intervention trials.

Additional research tests the robustness of the social-belonging intervention in mitigating other cases of underachievement and processes that contribute to its effectiveness. One study found that both the social-belonging intervention and a second, novel intervention raised women’s GPA in the first year of a demanding engineering program and yielded evidence that different psychological pathways mediated the two interventions (Walton, Logel, Peach, Spencer, & Zanna, in prep). Another found that the social-belonging intervention improved Black students’ transition to middle school (Walton, Cohen, Cook, Garcia, in prep), reducing disciplinary incidents among boys and improving grades, effects that unfolded over 3 years. Additional research with Dave Paunesku tests the effectiveness of delivering psychological interventions online (see www.perts.net). By reducing the marginal cost of additional participants, online delivery could provide a powerful vehicle to bring effective interventions to more at-risk students and thus afford large-scale trials and tests of contextual moderators (e.g., the effects of school demographics).

Questions of Mechanism: Understanding Effects Over Time

A critical question involves how brief psychological interventions generate long-lasting effects. One line of research explores the role of learning. Whereas past research finds that stereotype threat undermines intellectual performance in the immediate testing environment, Valerie Jones Taylor and I found that stereotype threat can also undermine learning (Taylor & Walton, 2011). In this research, Black students who studied academic material in a threatening learning environment scored worse on a test of that material a week later even when they performed in a nonthreatening setting. This finding shows that psychological threat can undermine future performance by preventing students from learning material in the first place. Conversely, removing threat can improve future performance. The results illustrate how interventions that remove psychological barriers can facilitate academic performance.

Another mechanism involves social-relational processes. Perhaps the social-belonging intervention helps minority-group students develop better relationships with majority-group members in school, and these relationships protect students against ongoing psychological threats. To examine this, in laboratory studies I have examined how interactions and relationships with majority-group members affect stereotyped students’ experience of threat. In one line of studies led by Christine Logel, boorish nonverbal behavior from men caused female engineers to experience stereotype threat and to perform worse on an engineering test (Logel, Walton, Spencer, Iserman, von Hippel, & Bell, 2009). Other research finds that positive relational behaviors, such as a professional handshake from a man to a woman, can mitigate stereotype threat (Akcinar, Cortes, & Walton, in prep; Carr, Walton, & Dweck, in prep). To examine whether better relationships sustain intervention effects over time, ongoing research tests whether interventions aimed at encouraging teachers to develop better relationships with students in the transition to high school improve minority-group students’ outcomes (with Christine Logel, Jason Okonofua, and Dave Paunesku) and whether intervention effects are mediated by change in students’ social relationships and peer networks (with David Yeager). This research affords insight into the interplay between psychological and system and social network processes.

In addition to these specific research questions, the long-lasting effects of social-psychological interventions raise broad questions for the field. Historically social psychology has emphasized the power of the immediate situation in shaping psychology and behavior. But if the proximate situation is all-powerful, how could brief interventions generate long-lasting effects? This issue raises important questions about how social-psychological processes unfold over time, which, I believe, may form a new frontier for social-psychological research, connect the field to other disciplines, and suggest unique remedies to social problems. I am excited to explore this broadly in future research.

Latent Ability: Implications for the Interpretation of Group Differences and for Affirmative Action

If widely-known negative stereotypes systematically undermine the academic performance of women and ethnic minority students, this would suggest a new interpretation of group differences in academic performance. Traditionally, group differences have been assumed to reflect real differences in ability. But if stereotypes are in play, then grades and test scores may underestimate the ability of women and ethnic minorities. A part of their ability would be “latent” or hidden by psychological threats.

With Steve Spencer, I tested this latent ability hypothesis in two novel meta-analyses, which combined data from 18,976 students in 5 countries in North America and Europe (Walton & Spencer, 2009). The meta-analyses found that, when psychological threat was removed from both laboratory and real-world settings, women and minorities performed better than men and non-minorities who had the same prior test scores and grades. The magnitude of this superior performance—called the latent ability effect—indexes the degree to which the prior measures underestimated the true ability of ethnic minorities and women. The size of the effect suggests that most of the gender gap on the SAT-Math test, for instance, and a significant portion of race gaps on the SAT are due to psychological threat.

These results carry important policy implications. They suggest (1) that negatively stereotyped students have a significant amount of hidden intellectual potential, which organizations can tap by removing psychological threats from their internal environments and (2) that in evaluating candidates for admission or hiring, interpreting grades and test scores at face value will discriminate against women and ethnic minority students if these measures were assessed in typical threatening environments. Because of this, my colleagues and I argue that organizations should follow specific procedures to account for latent ability in interpreting performance measures. In doing so, organizations could promote both diversity and meritocracy at once; we call this policy affirmative meritocracy (Walton, Spencer, Erman, 2013; see also Logel, Walton, Peach, Spencer, & Zanna, 2012).

The Social Self

As noted, a major theoretical emphasis of my research is that social belonging plays a broader and more important role in people’s functioning and psychology in general than is perhaps now understood. In a series of projects I have explored the hypothesis that a primary source of people’s sense of self lies in their social identity—their network of personal relationships and group identities.

Mere Belonging

One line of research finds that a minimal sense of social connectedness, even with unfamiliar others, can cause people to merge psychologically with others—for instance, to pursue the same interests and to feel the same emotions. This research examines what my colleagues and I call mere belonging, a state triggered by small cues of social connectedness to another person or group. One line of studies found that cues like a shared birthday with a math major or membership in a minimal “numbers group,” increased participants’ motivation in math, for instance increasing persistence on a challenging math puzzle (Walton, Cohen, Cwir, & Spencer, 2012). Related research found that mere belonging causes similar increases in motivation among young children, increasing motivation and learning among 4- and 5-year-olds (Master & Walton, in press). A third line of research found that mere belonging also causes people to experience shared emotions and physiological reactions (Cwir, Carr, Walton, & Spencer, 2011). In one study, after a confederate had jogged in place, participants who had been led to feel socially connected to the confederate had greater cardiovascular reactivity than participants who were merely exposed to the confederate. Many of these effects were mediated by participants’ sense of social connectedness to the peer at hand. Together this research shows that people acquire socially-important aspects of self vicariously as a consequence of cues of social connectedness; the boundaries between self and other are surprisingly fluid, so that the goals, motivation, and emotions of others readily become our own.

Related research shows implications for intergroup relations. Feeling minimally connected to a peer of another ethnicity led participants to exhibit increased interest in that person’s culture (Brannon & Walton, in press). Moreover, in studies drawing on psychology’s consistency theories (e.g., dissonance), after people freely acted on this interest by taking part in a cultural activity, which could seem inconsistent with holding prejudicial attitudes, they showed lower levels of prejudice and, in an unrelated context 6 months later, more positive intergroup attitudes, for instance reporting greater interest in talking with Mexican American peers about sensitive topics (e.g., immigration).

A Sense of Togetherness

Other research examines not a sense of personal social connection but a sense of pursuing a task or goal jointly with others. The capacity to work together to design and build tools, objects, and cultural products is a foundation of human culture that affords enormous social and personal benefits. These benefits illustrate the value a sensitivity to the psychological experience of working together would have, and suggest, Priyanka Carr and I hypothesized, that a sense of working together may fuel intrinsic motivation. Consistent with this hypothesis, laboratory studies found that people led through subtle means to feel that they worked with others rather than in parallel to others evidenced greater intrinsic motivation for a task, freely persisting longer, becoming more absorbed, and performing better (Carr & Walton, under review). Similar effects appear even among 4- and 5-year-olds (Butler & Walton, in prep).

This research suggests a new interpretation for the effects of social norms. Classic research suggests that people behave in line with norms because norms communicate what behaviors are effective and/or socially approved. Priyanka Carr and I hypothesized that, in addition, following norms can create a feeling of joining with others in pursuit of a shared goal. To test this hypothesis, in several studies we held constant normative information but manipulated whether this information created a feeling of joining together or did not. As predicted, when the feeling of joining together was evoked, participants followed the norm more; in one study, reducing paper-towel use in public restrooms more (Carr & Walton, under review). Ongoing research examines the effectiveness of a “norm + togetherness” intervention in reducing residential water and energy use in large-scale trials).

The Social Malleability of the Self

In addition to the effects of social belonging on the self, I am interested in how the self is socially constructed through top-down processes. One line of research examined subtle linguistic cues (Walton & Banaji, 2004). For instance, people viewed their preferences as stronger and more stable when they had been led to describe them using essentialist, noun-based language (e.g., “I am a chocolate-eater”) than descriptive action verbs (e.g., “I eat chocolate a lot”). Other work examines effects on behavior. With Chris Bryan, I hypothesized that nouns signal that a future behavior represents an identity—the kind of person one is or could become. When this identity is valued, we hypothesized, nouns should increase the likelihood that people will engage in this behavior. In one test of this hypothesis, participants completed one of two surveys the day before major elections. The survey questions featured either nouns (e.g., “How important is it to you to be a voter in tomorrow’s election?”) or descriptive action verbs (e.g., “How important is it to you to vote in tomorrow’s election?”). As assessed by state records, the noun condition caused an 11 percentage point increase in voter turnout (Bryan, Walton, Rogers, & Dweck, 2011). This is one of the largest boosts in voter turnout ever observed in a randomized trial, and could easily reverse the outcome of many close elections. The results highlight how people’s desire to see themselves as having positive identities can be channeled, even through subtle means, to motivate socially important behavior.

Another line of research examines the popular theory advanced by Baumeister and colleagues that people have a limited degree of willpower, which is easily depleted. Past research finds that after people exert self-control they perform less well on subsequent self-control tasks. This is commonly interpreted as evidence of “ego depletion.” Contrary to this notion, in a line of research led by Veronika Job, I explored the hypothesis that it is people’s theories about willpower that are responsible for “depletion” effects. Indeed, we find that it is only when people believe or are led to believe that willpower is limited (i.e., easily depletable) that they show depletion effects (Job, Dweck, & Walton, 2010). By contrast, a non-limited theory leads people to exert greater self-control (see also Miller, Walton, Dweck, Job, Trzesniewski, & McClure, 2012). Further a new intervention finds that teaching people a non-limited theory of willpower improves self-regulation in times of stress in real-world settings (e.g., reduced procrastination, Bernecker, Job, Dweck, & Walton, in prep).

A final line of research developed with Carol Dweck and with Dave Paunesku explores the functioning of the self. As in research on working selves (Markus & Wurf, 1987), we suggest that “selves” can be local but have powerful motivational effects. For instance, people may be motivated to view a currently activated self positively. Consistent with this hypothesis, we find that priming a McDonald’s “self” (e.g., by viewing a first-person slideshow of a walk into rather than past a McDonald’s) leads people to underestimate the caloric content of McDonald’s food. Activating an “airline passenger” self leads people to adopt a defensive view of airline safety (Paunesku, Walton, & Dweck, under review). Drawing on this reasoning, we suggest that one reason stereotype threat can have negative effects is because it defines the working self narrowly and negatively—in the spotlight of a stereotype, it may seem that nothing but the threatened group identity is relevant. Conversely, remedies to such threats often broaden the self, reminding people that there is more to them than the threatened aspect of self (Walton, Paunesku, & Dweck, 2011).

Want to learn more?

Read a dialogue with Philosophy Professor Michael Bratman here Thinking twice-Motivation

Want to learn more? Watch this  2-minute video or this 6-minute video describing the belonging intervention

Want to learn more? Read this op-ed on willpower, or check out this radio program.

Gregory M. Walton

  1. -Assistant Professor

  2. -Department of Psychology

  3. -Stanford University

Contact Information

  1. -(650) 498-4284

  2. -gwalton(at)stanford.edu

Curriculum Vita