2017-2018 Fall

Amanda McLean and Jon Krosnick

September 25, 2017

Accuracy of 2016 Pre-election Polls: More Success of Random Sampling

ABSTRACT: An analysis of pre-election polls conducted the week before the 2016 U.S. presidential election shows that probability samples were highly accurate, and river sampling surveys were highly inaccurate.

Amanda McLean

Lauren Howe, PhD Candidate, Department of Psychology

October 2, 2017

Normative Appeals Are More Effective When They Invite People to Work Together Toward a Common Cause 

ABSTRACT: Five experiments tested the hypothesis that appeals to social norms would inspire conformity more when they represent social norms as an opportunity to join with others to work toward a common goal than when they do not. Working-together normative appeals, which invited people to “join in” and “do it together,” increased interest in donating to charities (Experiments 1 and 4); increased actual charitable donations (Experiment 2); reduced paper towel use in public restrooms in an intervention field experiment (Experiment 3); and increased interest in reducing personal carbon emissions (Experiment 5). By contrast, appeals that included normative information alone did not affect interest or behavior. Mediation analyses suggest the effectiveness of working-together normative appeals arises because they simultaneously foster feelings of working together and free choice in addressing a common problem, while normative-information appeals create the feeling of being pressured, undermining their behavioral effects.

Lauren Howe
David Brookman, PhD Candidate, Graduate School of Business

October 9, 2016

Issue Publics and Research Design

ABSTRACT: In recent decades, an accumulating body of research has explored the role of attitude importance in generating “issue publics”, groups of people who feel strongly about an issue and are especially likely to use the issue in their thinking and action in their arena of politics. This presentation will describe a series of new studies we are planning to conduct to get feedback on design and interpretation.

David Broockman
John Protzko, PhD Candidate, University of California, Santa Barbara

October 16, 2017

Social Judgments & Metascience in 2017

John Protzko
Bo MacInnis, PhD Candidate, Department of Communication

October 23, 2017

Political Parties’ Economic Performance Affects Americans’ Attitudes, Emotions, and Behavioral Intentions

ABSTRACT: For decades, research on retrospective voting in the U.S. has suggested that citizens are more likely to vote for the incumbent party when the economy is good or improving than when it is in bad shape or declining. However, most such research is observational, cross-sectional associations between variables, time-series analysis, or lagged analysis using panel survey data. This study complements that work by offering strong tests of causal influence through experiments embedded in online national surveys of American adults. In a series of such experiments, respondents were randomly assigned to watch a video unrelated to the economy or an educational video showing that, since 1948, an economic indicator (either the annual unemployment rate or the annual average income tax rate) declined during presidencies of one major political party in the U.S. and rose during presidencies of the other major party.

Exposure to the educational video increased participants’ preference for the party portrayed as presiding over good economic developments, improved assessments of the party’s economic competence and its ability to solve the country’s problems, elevated positive affect toward and strengthened the identification with that party, and increased intention to vote for that party’s candidates. Thus, these studies demonstrated the effectiveness of a new type of educational political advertising:

advertising political parties instead of candidates, and in doing so, generated empirical support for retrospective voting theory’s causal claims.

Bo MacInnis
Jeremy Freese, PhD Candidate, Department of Sociology

October 30, 2017

Reproducibility and Transparency in the Social Sciences

ABSTRACT: Across the medical and social sciences, new discussions about replication have led to transformations in research practice. Sociologists, however, have been largely absent from these discussions. The goals of this review are to introduce sociologists to these developments, synthesize insights from science studies about replication in general, and detail the specific issues regarding replication that occur in sociology. The first half of the article argues that a sociologically sophisticated understanding of replication must address both the ways that replication rules and conventions evolved within an epistemic culture and how those cultures are shaped by specific research challenges. The second half outlines the four main dimensions of replicability in quantitative sociology—verifiability, robustness, repeatability, and generalizability—and discusses the specific ambiguities of interpretation that can arise in each. We conclude by advocating some commonsense changes to promote replication while acknowledging the epistemic diversity of our field.

Jeremy Freese
Alice Kathmandu, School of Education

November 6, 2017

Voting Machine and the Name Order Effect

ABSTRACT: Voting machines, such as punch cards, optical scan cards, and touchscreens, are currently seen only as conduits between a voter’s intention and a cast on the ballot. Little is known about the voting machine effect. We hypothesize that using punch cards and optical scan cards involve bodily motion and physical change on an immediate object, and would thus induce more deliberation among voters, compared with using touchscreens. We test this hypothesis with the 2004 Presidential Election data from Ohio. Initial evidence suggests that both punch cards and optical scan cards facilitate voter deliberation and reduce cognitive bias, whereas touchscreen generates the most biased results. These findings call into question the previously unquestioned role of voting machines as the cornerstone of our democracy: some machines induce voter bias more than others.

Catherine Heaney, PhD Candidate, School of Medicine

November 13, 2017

The Stanford WELL Project 

ABSTRACT: The Stanford Prevention Research Center is conducting an international multi-site study of the nature of well-being and explorations of how best to positively influence well-being. In this talk, I will discuss the process we used to develop a strong conceptualization of well-being and a valid and reliable survey measure. The complexities of conducting cross-cultural research in this arena will be described.

Catherine Heaney
Bonnie Halpern-Felsher, PhD Candidate, School of Education

November 27, 2017

Adolescents’ and Young Adults’ Perceptions of Health and Social Risks and Marketing across Tobacco Products

ABSTRACT: While adolescent cigarette use has declined, overall use of tobacco and nicotine has remained steady or increased, owing largely to e-cigarette, hookah, smokeless tobacco, and little cigar/cigarillo use. Numerous studies have examined the influence of mis(perceptions) of risks and benefits, product packaging and marketing on cigarette use. However, few studies have examined these factors across tobacco products. Dr. Halpern-Felsher’s research examines cognitive, psychosocial and environmental influences on adolescents’ and young adults’ tobacco use, with a particular emphasis on their perceptions of specific social and health risks, addiction risks, and benefits; and how exposure to marketing influences these perceptions and subsequent tobacco use. She will discuss an NIH-funded prospective longitudinal study examining adolescents’ and young adults’ perceptions across tobacco products and marijuana, and how perceptions predict use. Dr. Halpern-Felsher’s talk will focus on perceptions of risks and benefits across tobacco products, perceptions of addiction, attitudes towards e-cigarettes and e-cigarette users, perceptions of product packaging, and perceptions of marketing of e-cigarette flavors and cessation aids. Her study results inform health education, policy, and regulation.

Bonnie Halpern-Felsher
Mira Lindner, Department of Political Science, Aarhus University

December 4, 2017

Does Coalitional Affiliation Override Race in Shaping Punishment of Terror-Suspects? An Experimental Study

ABSTRACT: Previous research has shown that outgroup offenders are punished more harshly than ingroup offenders, both in domains of criminal justice and terrorist violence. Because terrorism is a group-based phenomenon in which race and ethnicity are particularly salient, most studies have focused on in- and outgroup membership along these lines. Recent psychological research, however, suggest that cooperative alliances – or coalitions – can override categorization by race. In times where an increasing number of terrorist attacks are perpetrated by home-grown terrorists, this poses an interesting question: To what extent is punishment a function of a terrorist’s adherence to a specific coalitional ideology, rather than their race or ethnicity? In order to tackle this question, we will conduct an experiment on a nationally representative sample of 2,000 U.S. adults. Specifically, we examine the interaction of race/ethnicity of the perpetrator (White vs. Black) and adherence to a specific terrorist organization (domestic vs. international) on proposed severity of punishment. We expect that Whites are punished less severely for their offense than Blacks. This association, however, should be contingent upon affiliation with a terrorist organization – specifically, we expect punishment of Whites to approach that of Blacks for those adhering to an international terrorist organization. A confirmation of these hypotheses would suggest that affiliation with a specific terrorist ideology can override categorizations based on race, bearing important ramifications for media discourse after terrorism-related emergencies and mass shootings.