Figure 1.
Experiment 1 preferences for which study to conduct (1=strong preference for the gun control/skin cream ineffective study, 3=no preference, 5=strong preference for the gun control/skin cream effective study) as a function of ideology (liberal, moderate, conservative), polarization condition (experimental vs. control), and prior evidence (absence of evidence, presence of conflicting evidence).
Error bars indicate one standard error above and below the mean. Dotted line indicates neutral preference.
