| "Attitude polarization" research has received great attention, yet has typically focused on self-reports of (self-perceived) change. When it has not, it has used general attitude measures (e.g., death penalty support), typically in a before and after manner. How much polarization might actually affect particular case-based judgments has been unclear.; In this study, participants (classified as having beliefs biased toward punishment or toward rehabilitation using an indirect method) made judgments regarding a particular crime scenario. Those receiving pro-punishment or pro-rehabilitation manipulations (or a mixture of both), were compared to similarly belief-classified participants who made such judgments in a view-neutral context.; Evidence for a general movement toward a presented pro-punishment or pro-rehabilitation view was only found for a central (neutral) segment of the belief spectrum used to predict punishment-bias. In contrast, participants whose beliefs indicated initial bias toward agreeing with punishment or rehabilitation views only showed strengthening of their positions; that is, "polarization" characterized these initially-biased participants. Results indicate that polarization (of deserved punishment judgments) occurs not only from encountering mixed belief-supporting and belief-opposing information (as previous research suggested) but by either information alone.; While the pro-punishment and pro-rehabilitation manipulations notably affected subjective "deserved punishment" judgments, anger (and disgust), and (to a lesser extent) sentence recommendations, manipulations did not generally affect rehabilitation judgments or associated emotions of sadness, hope, or sympathy; these instead seemed dependent only on initial beliefs. Punishment and rehabilitation judgments, along with associated emotions (anger and sympathy/sadness/hope, respectively) showed fairly strong and unique contributions to sentence recommendation predictions. Marginally significant polarization effects (due to receiving views opposing participants' beliefs) were found for sentence recommendations, suggesting that "polarization" may be a phenomenon that extends beyond self-reports, general attitude measures, and even beyond subjective punishment judgments. Belief-classification did not predict differences in perceived likelihoods of either possible verdict error, but instead predicted estimates of: how lazy and selfish typical criminals are, the percentage of truly guilty defendants, and the relative frequencies of verdict errors.; This study has implications for the relationships among punishment, rehabilitation, anger, sadness and sympathy, as well as for how sentencing instructions or defense/prosecution arguments affect jurors' decisions. |
