The aim of this research was to test an informational explanation of the effects observed in the affect misattribution procedure (AMP). According to this explanation, participants performing the AMP would simplify the task by asking whether the target is pleasant (yes vs. no) and would use the affective information provided by the prime to answer the question (positive = yes, negative = no). In line with this proposition, we observed in three preregistered experiments that slightly modifying the response options proposed in the task moderated the effect, which can be canceled (Experiment 1) and even reversed (Experiments 2 and 3). These results are consistent with the informational explanation and seem difficult to explain by the operation of misattribution processes.