1,721,011 research outputs found
Persuasion in Parallel
No full textAccording to theories of motivated reasoning, attempts to persuade political opponents are often counterproductive because they end up strengthening opponents’ initial views via directional motivations. Drawing on evidence from 23 randomized survey experiments, Persuasion in Parallel shows that the predicted “backlash” fails to materialize. Instead, the experiments show that the effects of persuasive messages are similar for many subdivisions of society, including policy opponents and proponents, Republicans and Democrats, young and old, and men and women. The overarching conclusion is that persuasion occurs in parallel: even though Americans differ tremendously in their baseline views on many political issues, they are quite similar in their responses to information. This empirical pattern casts serious doubt on the motivated reasoning framework for understanding information processing. The political implication of this work is that we should not give up trying to persuade the other side
Replication Data for: The Long-lasting Effects of Newspaper Op-Eds on Public Opinion
No full textDo newspaper opinion pieces change the minds of those who read them? We conduct two randomized panel survey experiments on elite and mass convenience samples to estimate the effects of five op-eds on policy attitudes. We find very large average treatment effects on target issues, equivalent to shifts of approximately 0.5 scale points on a 7-point scale, that persist for at least one month. We find very small and insignificant average treatment effects on non-target issues, suggesting that our subjects read, understood, and were persuaded by the arguments presented in these op-eds. We find limited evidence of treatment effect heterogeneity by party identification: Democrats, Republicans, and independents all ap- pear to move in the predicted direction by similar magnitudes. We conduct this study on both a sample of Amazon Mechanical Turk workers and a sample of elites. Despite large differences in demographics and initial political beliefs, we find that op-eds were persuasive to both the mass public and elites, but marginally more persuasive among the mass public. Our findings add to the growing body of evidence of the everyday nature of persuasion
Replication Data for: Do Belief Systems Exhibit Dynamic Constraint?
No full textAs described in \citet{Converse:1964}, belief systems are dynamically constrained if a change in one opinion causes a concomitant change in a related opinion. While an enormous literature is dedicated to the study of static constraint (the extent to which individuals hold political views that ``go together''), dynamic constraint is rarely studied, especially using experimental research designs. We offer a new formalization of the theoretical argument that suggests an identification strategy for detecting dynamic constraint. We present evidence from survey experiments conducted with convenience samples of both the mass public and of political elites. Our results indicate that even among respondents whose belief systems are highly constrained in the static sense, a change in one attitude need not precipitate changes in related attitudes. These experimental results affirm and extend Converse's thesis about the limited extent of dynamically constrained ideological thinking in the mass public. The lack of dynamic constraint among our elite sample raises the question of how they come to hold political opinions that are constrained in a static sense. We present an experiment that suggests a potential explanation: elites may be more likely to be chided for expressing inconsistent positions
Replication Materials for: Visualize as You Randomize: Design-Based Statistical Graphs for Randomized Experiments
No full textA good statistical graph for a randomized experiment simultaneously conveys the study's design, analysis, and results. It reveals the experimental design by mapping design elements to aesthetic parameters. It illuminates the analysis by plotting the statistical model in ``data-space.'' When the design and analysis of an experiment are encoded in a plot, the interpretation of the experimental results is clarified. ``Analyze as you randomize'' is a dictum attributed to Fisher that guides interpretations of experimental data. This chapter extends that principle to visualizations of randomized experiments. While not every experiment requires a visualization, those that do should be visualized in ways that communicate the design and results together
Replication Data for: Generalizing from Survey Experiments Conducted on Mechanical Turk: A Replication Approach
No full textTo what extent do survey experimental treatment effect estimates generalize to other populations and contexts? Survey experiments conducted on convenience samples have often been criticized on the grounds that subjects are sufficiently different from the public at large to render the results of such experiments uninformative more broadly. In the presence of moderate treatment effect heterogeneity, however, such concerns may be allayed. I provide evidence from a series of 15 replication experiments that results derived from convenience samples like Amazon's Mechanical Turk are similar to those obtained from national samples. These results suggest that either the treatments deployed in these experiments cause similar responses for many subject types or convenience and national samples do not differ much with respect to treatment effect moderators. Using evidence of limited within-experiment heterogeneity, I show that the former is likely to be the case. Despite a wide diversity of background characteristics across samples, the effects uncovered in these experiments appear to be relatively homogeneous
Replication Data for: The Long-lasting Effects of Newspaper Op-Eds on Public Opinion
No full textDo newspaper opinion pieces change the minds of those who read them? We conduct two randomized panel survey experiments on elite and mass convenience samples to estimate the effects of five op-eds on policy attitudes. We find very large average treatment effects on target issues, equivalent to shifts of approximately 0.5 scale points on a 7-point scale, that persist for at least one month. We find very small and insignificant average treatment effects on non-target issues, suggesting that our subjects read, understood, and were persuaded by the arguments presented in these op-eds. We find limited evidence of treatment effect heterogeneity by party identification: Democrats, Republicans, and independents all ap- pear to move in the predicted direction by similar magnitudes. We conduct this study on both a sample of Amazon Mechanical Turk workers and a sample of elites. Despite large differences in demographics and initial political beliefs, we find that op-eds were persuasive to both the mass public and elites, but marginally more persuasive among the mass public. Our findings add to the growing body of evidence of the everyday nature of persuasion
Replication Materials for: Visualize as You Randomize: Design-Based Statistical Graphs for Randomized Experiments
No full textA good statistical graph for a randomized experiment simultaneously conveys the study's design, analysis, and results. It reveals the experimental design by mapping design elements to aesthetic parameters. It illuminates the analysis by plotting the statistical model in ``data-space.'' When the design and analysis of an experiment are encoded in a plot, the interpretation of the experimental results is clarified. ``Analyze as you randomize'' is a dictum attributed to Fisher that guides interpretations of experimental data. This chapter extends that principle to visualizations of randomized experiments. While not every experiment requires a visualization, those that do should be visualized in ways that communicate the design and results together
Replication Data for: Avoiding Post-Treatment Bias in Audit Experiments
No full textThis archive contains the data original posted by White et al. here: http://dx.doi.org/10.7910/DVN/28158 and a reanalysis of those data
Replication Data for: Generalizing from Survey Experiments Conducted on Mechanical Turk: A Replication Approach
No full textTo what extent do survey experimental treatment effect estimates generalize to other populations and contexts? Survey experiments conducted on convenience samples have often been criticized on the grounds that subjects are sufficiently different from the public at large to render the results of such experiments uninformative more broadly. In the presence of moderate treatment effect heterogeneity, however, such concerns may be allayed. I provide evidence from a series of 15 replication experiments that results derived from convenience samples like Amazon's Mechanical Turk are similar to those obtained from national samples. These results suggest that either the treatments deployed in these experiments cause similar responses for many subject types or convenience and national samples do not differ much with respect to treatment effect moderators. Using evidence of limited within-experiment heterogeneity, I show that the former is likely to be the case. Despite a wide diversity of background characteristics across samples, the effects uncovered in these experiments appear to be relatively homogeneous
Replication Data for: Avoiding Post-Treatment Bias in Audit Experiments
No full textThis archive contains the data original posted by White et al. here: http://dx.doi.org/10.7910/DVN/28158 and a reanalysis of those data
- …
