Purpose: The aim of the study was to determine if memory for words spoken in noise depends on vocal emotion (fear, neutral, pleasant surprise, or sadness) to a greater extent if the listeners are engaged in an emotion identification task (current experiment) compared to only repeating the words (previous experiment).

Hypotheses: It was hypothesized that there would be an effect of emotion on repetition and recall and that this effect would be greater for repetition and recall accuracy when listeners were engaged in a task that drew attention to the identification of the vocal emotion. Performance was expected to be better for words portraying arousing emotions (fear and pleasant surprise) than for words spoken with sad or neutral emotion.

Methods: Participants listened to 100 sentences spoken in four different emotion conditions. All words were semantically neutral. Participants were instructed to 1) repeat the word presented, 2) identify the emotion in which it was spoken, 3) judge whether the word began with the first or second half of the alphabet, and 4) after each set, recall as many words as possible that had been heard in the set.

Results: Repetition accuracy was higher for words spoken to portray fear than sadness, but pleasant surprise was the same as neutral. Similarly, recall accuracy was highest for words spoken to portray fear, and lowest for sadness. The addition of the emotion identification task did not alter performance. Emotion identification was highest for pleasant surprise and lowest for neutral.