Accepted at Political Behavior

Recent research on political persuasion suggests that narrative- and evidence-based messages can effectively change attitudes in settings such as canvassing, video presentations, and survey experiments. Yet less is known about how these persuasive strategies operate in free-flowing, two-way, cross-partisan political conversations. This study examines the relationship between persuasive message types, conversational reciprocity, and post-discussion attitudes using survey and text data from 966 U.S. participants engaged in anonymous one-on-one political discussions on DiscussIt, a mobile chat platform. I analyze how personal narratives and evidence-based arguments relate to conversational interaction outcomes and measurable attitude change. Participants exposed to more personal narratives were more likely to perceive their conversation partner as influential and report more positive conversational experiences, particularly when narrative exchanges were mutually reciprocated. In contrast, evidence-based exchanges were associated with lower perceived interpersonal influence, lower conversational enjoyment, and greater perceptions of relative knowledge. Despite these differences in conversational dynamics, neither persuasive strategy across conversation structures was associated with significant changes in immediate or delayed political attitudes. Unlike studies of unidirectional persuasive messages, this research examines spontaneous peer-to-peer political conversations. Results suggest that persuasive strategies influence relational and experiential outcomes more than political attitudes.

2024 James W. Protho Student Paper Award Honorable Mention

2026. American Political Science Review (with Trent Ollerenshaw and Andrew Trexler)

We re-examine recent influential claims that repeated measure experimental designs do not introduce bias and offer large precision gains in survey research (Clifford, Sheagley, and Piston 2021). We test these claims by experimentally varying the design of six classic political science experiments across three distinct large samples of U.S. adults (total N = 13, 163). In contrast to the original study, we observe consistent attenuation of treatment effects in repeated measure designs. However, this average design effect is small enough, and the precision gains large enough, that largely we affirm the recommendation to employ repeated measure designs in many practical research applications—though post-only designs with large samples may be preferable when estimating the precise magnitude of the treatment effect is a priority. We additionally extend the literature on repeated measure designs by exploring how several design considerations affect the bias-precision trade-off, such as the use of within-subject versus between-groups designs, the relative separation of repeated measures within single surveys, and differences in respondent characteristics across sample types.

2025. Journal of Personality and Social Psychology (with Sanaz Talaifar, Gabriella Harari, and Samuel Gosling)

Socializing, moving, working, and leisure form the foundation of human experience. We examined whether these foundational, ostensibly nonpolitical behaviors are nevertheless bifurcated along political fault lines, revealing “lifestyle polarization.” Study 1 quantified the association between political identity and 61 social, movement, work, and leisure behaviors collected from smartphone sensors and logs (i.e., GPS, microphone, calls, texs, unlocks, activity recognition) and ecological momentary assessments (i.e., querying activity level, activity type, interaction partners, locations) at multiple temporal levels (i.e., daily, mornings, afternoon, evenings, nights, weekends, weekdays) in a sample of up to 1,229 students on a college campus. We found that liberals and conservatives behave differently in everyday life; the behavioral differences were small but robust, not accounted for by other plausible factors (e.g., demographics), and most pronounced in the leisure domain. Study 2 showed that the behavioral differences between liberals and conservatives were not accurately discerned by other students, who overestimated the extent of lifestyle polarization present on their campus. Together, these studies suggest that political identity has penetrated some of the most foundational aspects of everyday life, but not to the degree that people think. We discuss how communities may feel divided not only because of deep ideological disagreements across partisan lines but also because such disagreements are accompanied by distinct lifestyles—both real and (mis)perceived—that may prevent liberals and conservatives from engaging in cross-partisan contact and developing mutual understanding.

Social media creates opportunities for political discussion across partisan divides, yet political interactions online are often characterized by incivility and disengagement. Generative AI may offer a scalable approach to fostering more constructive political communication, but it remains unclear whether AI assistance can improve political interactions in multi-user social media environments. We investigate this question through a preregistered experiment with 1,194 participants recruited from a probability-based sample. Participants were randomly assigned to a customized social media environment with either no AI assistant or an AI assistant that provided conversational nudges encouraging respectful communication, engagement with opposing viewpoints, and acknowledgment of alternative perspectives. We find that AI assistance did not improve participants’ perceptions of dialogue productivity, evaluations of other users, or evaluations of the platform. However, it changed observable patterns of engagement. Participants assigned to an AI-assisted feed commented more frequently, were more likely to comment on posts expressing opposing viewpoints, and were more likely to make polite comments on those posts. These findings suggest that AI assistance can shape how people engage with political disagreement, even when it does not alter their perceptions of those interactions or their broader orientation toward disagreement.

The diploma divide has become a defining feature of contemporary American politics, with college graduates increasingly aligning with the Democratic Party, while Americans without college degrees are increasingly supporting the Republican Party. Yet college graduates are not politically homogeneous, and first-generation status may reveal important departures from this pattern by capturing differences in social origins. Using General Social Survey data from 2000–2024, I compare first-generation and continuing-generation graduates across partisanship, presidential vote choice, ideology, and political, economic, and social attitudes. I find that first-generation graduates are more likely than continuing-generation graduates to identify as Republican, vote Republican, and hold conservative views, while remaining less conservative than non-college Americans. These differences persist after accounting for demographic characteristics, childhood economic circumstances, and current socioeconomic conditions. First-generation graduates also express more conservative economic, social, and cultural attitudes than continuing-generation graduates. Taken together, the findings suggest that first-generation college graduates occupy a distinctive position within the political landscape of the college-educated population. Their attitudes do not consistently mirror those of continuing-generation graduates or non-college Americans, indicating that educational attainment alone does not fully capture the relationship between college education and political orientation. Instead, family educational background remains consequential for understanding political and social attitudes, suggesting that the diploma divide obscures important variations within the college-educated population.

Congressional communication occurs in a variety of venues that each have distinct affordances and audiences, which may make them differentially useful for discussing particular topics in particular ways. However, substantive research in congressional communication typically examines one venue at a time. We test the extent to which venue selection on the part of the researcher may contribute to the inferences they draw by exploring what topics members discuss and how they discuss them across Twitter, Facebook, press releases, e-newsletters, and one-minute speeches in the 116th Congress. We find that while some highly salient topics are discussed at similar rates across all venues, others receive differential emphasis and partisan differences in how topics are discussed also vary across venues. Our results suggest that researchers should explicitly consider how features of their data sources may encourage some forms of communication over others, and should make cross-platform comparisons when possible.

In prep

Liberals report lower wellbeing than conservatives. We provide the first evidence that this wellbeing gap can be partially explained by liberals’ greater social media use relative to conservatives. Ecological momentary assessments and smartphone app usage data from a university sample (N = 1,631 participants, k = 516,076 observations) and survey data from a nationally representative sample (N = 5,750 participants) revealed that liberals used social media more than conservatives did, which mediated the relationship between liberalism and lower wellbeing (i.e., life satisfaction, affective wellbeing, loneliness, sadness, anger, happiness, acceptance, worry, depression, anhedonia). Overall, social media use explained up to 15% of the ideological wellbeing gap. Effects were unique to social media use (e.g., TikTok); other kinds of smartphone use (e.g., for creativity, information) did not explain the wellbeing gap. Findings underscore the importance of disaggregating different types of digital media use and highlight the role everyday behavior in explaining liberal’s lower wellbeing.