Uncovering the relevance of reasons for behavior: The attitude-behavior gap revisited

To use a particular reason to explain behavior, the reason (e.g., to protect the environment) must be present when people engage in the action (e.g., riding a bike) and absent when people do not (e.g., not riding a bike). This thinking resonates in the statistical benchmark that behavioral scientists typically apply when assessing a reason's behavioral relevance. In contrast to what the notorious attitude-behavior gap insinuates, explaining small amounts of variance in a behavior does not inevitably challenge the behavioral relevance of reasons. The problem arises because different people have different reasons for engaging in a behavior and even for not engaging in it. By reanalyzing two previously collected data sets, we corroborate the environmental-protection reason's sensitivity for actions and specificity for inactions. Additionally, we confirm that both effects become even more convincing when person-specific rather than behavior-specific benchmarks for the presence and absence of a reason are employed.