When attitudes and habits don’t correspond: Self-control depletion increases persuasion but not behavior

Conversations with people who hold opposite partisan attitudes can elicit defensiveness, reinforce extreme attitudes, and undermine relationships with those with opposing views. However, this might not be the case when speakers experience high-quality (attentive, 2 understanding, and non-judgmental) listening from their conversation partners. We hypothesized that high-quality listening will increase speakers’ positive views toward, and their willingness to further interact with, others who hold politically opposed attitudes, and that these effects will be mediated by greater state openness. We conducted three experiments using different modalities to manipulate listening. In Study 1 (N = 379), participants recalled a conversation with an opposing political party member, with listening quality described as high-quality, low-quality, or control. Study 2 (N = 269) used imagined interactions, with participants reading vignettes describing either high-quality listening or a control condition. In Study 3 (preregistered; N = 741), participants watched a video of a listener modeling high-quality or moderate-quality listening and imagined themselves engaging in a similar interaction. Across studies, we found that high-quality listening consistently increased speakers’ state openness to politically opposed others, but did not change political attitudes. We found inconsistent evidence for speakers’ increased willingness to engage in future interactions (meta-analytic effect: 𝑑 = 0.20, p = .015). However, the indirect effect of listening on positive attitudes and willingness for future interactions through increased openness was consistently significant.

Four experiments were conducted to determine whether participants’ awareness of the performance criterion on which they were being evaluated results in higher scores on a criterion-valid situational interview (SI) where each question either contains or does not contain a dilemma. In the first experiment, there was no significant difference between those who were or were not informed of the performance criterion that the SI questions predicted. Experiment 2 replicated this finding. In each instance, the SI questions in these two experiments contained a dilemma. In a third experiment, participants were randomly assigned to a 2 (knowledge/no knowledge provided of the criterion) X 2 (SI dilemma/no dilemma) design. Knowledge of the criterion increased interview scores only when the questions did not contain a dilemma. The fourth experiment revealed that including a dilemma in a SI question attenuates the ATIC-SI relationship when participants must identify rather than be informed of the performance criterion that the SI has been developed to assess.