Emotional descriptors of words as predictors of recall in a paired-associate task

An extended replication of Contini and Whissell's 1992 experiment (N = 64 subjects) confirmed that there is a recall disadvantage for emotional words, especially active words, whether these serve as the cue or as the associate in a paired-associate task. A regression formula with trials, two gender variables, and seven emotional predictors (R2 = .77) was successful in predicting recall for 160 trials. Words or their CVC associates were better recalled if the words were emotionally neutral and least well recalled if the words were active or both unpleasant and passive.