What's recalled depends on the nature of the recall procedure: The case of mnemic neglect

According to the mnemic neglect model, people process non-threatening feedback more deeply than threatening feedback. Tests of the model rely on behavior recall as the primary dependent variable. Similar to other research programs in social psychology, little attention has been paid to determining the optimal recall procedure for testing the model and replicating its predicted findings. Four experiments reveal that the results of mnemic neglect studies are significantly affected by recall period length and how recalled behaviors are reported. A few basic principles (e.g., output interference, output primacy as an index of accessibility) can account for these findings. The lessons learned apply not just to mnemic neglect research, but to any investigation of social-cognitive processes utilizing free recall measures.