Research Interests
The goal of my research is to understand why persuasion attempts often fail and to clarify how prior attitudes, emotions, and social identities combine to guide attitude and behavior change. My work aims to address a variety of social issues, focusing predominantly on the unique theoretical and practical challenges for three issues: racial political divides, organizational diversity, and health disparities. In the first line of research, I examine how and why White Americans disengage with information about racism. A second line tests strategies for reducing resistance to DEI training and programs. Finally, a third line tests nuanced strategies for increasing receptivity to stigmatized health information. Across these lines, I advance theory and inform practice by integrating insights from diverse literatures (attitudes, intergroup processes, emotions) across disciplines (psychology, organizational behavior, communication, political science) using a variety of data (experimental, observational, and qualitative) and advanced analytic techniques (SEM, multilevel modeling).
How and why White Americans disengage with information and discussions about racism
To understand social problems, take them seriously, and support appropriate solutions, it is important that people have high-quality information about the topic. On the topic of racism, reading high-quality information is known to help correct pervasive misperceptions of racial inequality (e.g., Onyeador et al., 2019). Despite this benefit, concerted efforts to limit the dissemination of this kind of information are growing quickly, with many states working to ban discussions of racism in schools and private businesses. As conflicts over this kind of information grow, it is especially important to understand who is motivated to avoid information about racism and why.
​
My research finds that compared to their racially minoritized peers, White people prefer to avoid information about racism, even when it aligns with their racial and political attitudes (Takahashi, Jefferson, & Earl, in prep). Across eight studies, we found that perceived threats to White racial identity motivate White liberals and conservatives alike to avoid information about racism. We find direct evidence that White Americans simultaneously assess how unfavorable information may be to their political attitudes and to their racial identity. These assessments helped to explain complex patterns of information selection, whereby White conservatives and liberals alike were less likely to select attitude-consistent information if it simultaneously threatened their racial identity. Furthermore, I found that an intervention that encouraged people to seek discomfort as a necessary part of learning (Woolley & Fishbach, 2022) increased White Americans’ selection of information about racism. However, the intervention increased selections of different kinds of information for White conservatives and liberals. The intervention led White conservatives to select more information highlighting racism as a problem. For White liberals, however, the intervention led to increased selection of information denying racism as a problem. Together, this research suggests that White identity-based motivations are a barrier to engagement with the kinds of information that may improve public understandings of racism, and that messaging strategies that do not consider conflicts between identity and political attitudes risk backfiring.
​
Building on this research into the role of White identity threats, I am also investigating the specific contexts in which White Americans perceive their racial advantages and disadvantages, and what impact these nuanced perceptions have on their willingness to discuss racism. In one series of studies, I demonstrate that White Americans perceive the advantages and disadvantages of Whiteness in highly context-specific ways (Takahashi & Jefferson, in prep; preprint: https://psyarxiv.com/ry97q/). Although White participants on average perceived themselves as being advantaged in social and economic domains such as housing and education, they perceived themselves as distinctly disadvantaged in race discussions. I find that these context-specific feelings of disempowerment are shaped in part by perceptions that White people are silenced in race discussions—perceptions that lead to negative expectations for racial dialogue. Although perceiving racial silencing made White participants feel disrespected and devalued, it did not make them more motivated to defend their voice in these discussions. Instead, we found initial evidence that they were more likely to downplay the importance of discussing race. This points to a critical distinction between contexts in which privileged groups feel they are losing power and contexts in which they are unwilling to lose power.
Strategies for effective persuasion in DEI programs
A growing number of organizations are implementing training to address issues of diversity, equity, and inclusion. However, these efforts are often either insufficient or backfire altogether. Amidst calls for more comprehensive, evidence-based approaches to improving DEI, my research examines specific features of DEI programs that advance or impede concrete DEI goals.
​
In the context of DEI training, one important question is whether specific strategies commonly used to persuade audiences are indeed effective. For example, many DEI training programs present empirical research on the nature of bias and its impact within organizations. However, there is surprisingly limited evidence that these scientific explanations specifically—above and beyond other components of the training—help to explain the success of these programs. In the context of hiring, my colleagues and I have tested the features of a faculty recruitment workshop that successfully shifted hiring attitudes and behavioral intentions of university faculty on hiring committees (Sekaquaptewa, Takahashi, Malley, Herzog, & Bliss, 2019). I find that presenting faculty with social-scientific research on hiring biases increased their belief in unintentional biases in hiring, which helped to explain why the workshop increased endorsement of the recommended concrete strategies for reducing bias.
​
Other ongoing research investigates how structured racial dialogue programs are shaped by the racial composition of dialogues and the identity-based motivations people bring into the dialogues. Although structured intergroup dialogue programs have many proven benefits, these dialogues often unfold in ways that prioritize the education of privileged students, often at the frustration of marginalized students (e.g., Gallaway, 2017). Using mixed qualitative and quantitative methods, I am currently analyzing pre- and post-dialogue essays from students participating in the Intergroup Relations program at the University of Michigan over the span of three years. Specifically, I am investigating student essays from dialogues about race and racism composed of both students of color and White students, of students of color only, or of White students only. This work aims to understand students' initial goals for intergroup dialogue and how group context (i.e., intergroup vs. intragroup) and group status combine to shape progress toward these goals. Moving forward, my goal is to use insights from these research lines to develop DEI messaging strategies that can more precisely address the potentially conflicting needs and motivations of diverse audiences.
​
Similarly, I am also testing how various DEI messages from real-world organizations threaten or affirm the attitudes and identities of different audiences. Rather than changing the content of messages to be less threatening, this work then aims to increase receptivity to DEI messages by encouraging privileged group members to engage with messages they find threatening. This ongoing work reflects a key direction for my research moving forward, which is to test DEI messaging strategies that can more effectively and flexibly adapt to the conflicting identity-based and attitudinal motivations of heterogeneous audiences.
Nuanced strategies for increasing receptivity to stigmatized health information
A key insight that guides my research is that successful persuasion depends on many distinct processes that can conflict. This is a particularly big problem for health messages, as messaging strategies often produce mixed reactions that lead to mixed results. For example, people are often more persuaded by fear-inducing health messages (Tannenbaum et al., 2015) but tend to avoid engaging with them in the first place (Earl & Albarracín, 2007). Understanding the divergent effects an emotional state can have on different stages of the persuasion process will provide useful insights into effective health messaging.
​
In one paper, I tested new ways of using emotions to promote receptivity to health messages. Given that people tend to avoid messages that induce high-arousal negative emotions, such as fear or shame, a common approach for health messaging is to change the content of a message to reduce negative emotional reactions. However, this approach can only go so far if the topic itself elicits negative emotions, as is often the case with scary or stigmatized health conditions (e.g., HIV). A potentially useful but understudied approach would be to induce positive affective states before people even encounter potentially threatening information. Importantly, although the emotions known for disrupting attention are both negative and high arousal, it is unclear whether valence or arousal is likely to affect message processing, as prior work often conflates these dimensions. For this reason, I conducted a series of experiments (total n = 1,447) to investigate whether and how low-arousal positive affect impacts receptivity to health messages (Takahashi & Earl, 2020). I used a brief guided meditation (vs. a control audio task) to induce low-arousal positive affect before participants read health messages. I consistently found that the positive affect from meditation increased willingness to pay attention to health information—regardless of how threatening it was. Independently, the reduced arousal increased participants’ capacity to comprehend the message. Distinguishing the pathways through which valence and arousal impact willingness and capacity to process messages helps to resolve inconsistencies in past research, which generally examines high-arousal positive affect with mixed success. More importantly, this research demonstrates that affect can be leveraged to increase engagement with threatening health messages without changing the message content, although doing so requires careful attention to the specific part of the persuasive process targeted by an intervention.
​
My research is also focused on understanding how identity-based concerns can affect information processing for stigmatized group members. In the context of race and health messages, past research suggests that Black Americans value HIV information (i.e., positive attitudes about HIV information; Kaiser Family Foundation, 1998), but tend to pay less attention to it in clinical settings than White Americans (Earl et al., 2016). In one paper, my colleagues and I investigated how heightened concerns about HIV stigma among Black Americans disrupts Black patients’ attention to HIV information in real-world settings (Lewis, Kougias, Takahashi, & Earl, 2020 Health Communication). In a quasi-experimental field study in the waiting room of a public health clinic (n = 260), Black patients, but not White patients, paid less attention to HIV informational videos (but not flu videos) if there were other same-race patients in the room. This suggests that heightened stigma concerns from members of one’s own community helps to explain why Black patients may avoid stigmatized health information in clinical settings. Importantly, the behavior of other Black patients could also increase attention. Black patients paid more attention to the HIV video (compared to a control flu video) if other Black patients were visibly looking at the HIV video. These findings show that social identity and group membership are important for understanding receptivity to stigmatized health information in clinical settings, and that the behaviors of others who share one’s social identity can promote attention to otherwise threatening health messages.