Failure to replicate mood-dependent retrieval

We tried to replicate mood-dependent memory retrieval in a two-list interference design, as reported earlier in Experiment 3 of Bower, Monteiro, and Gilligan (1978, “Emotional mood as a context for learning and recall,” Journal of Verbal Learning and Verbal Behavior, 17, 573–585). Subjects in the experimental condition learned two word lists for free recall, one list while feeling happy, a second list while feeling sad. Later they recalled both lists while feeling happy or sad. Results revealed no mood dependency in recall; subjects recalled about the same percentage of words from the same-mood list as from the opposite-mood list. Thus, mood-dependent retrieval must be judged to be an unreliable phenomenon in the laboratory.