Learning the names of people: the role of image mediators.

Four experiments are reported involving the effects of bizarre and common imagery mediation techniques on the learning and 1-week retention of surnames, given videotaped faces as cues. The videotapes contained 24 undergraduates who were photographed from about the chest up, and who introduced themselves at a 20-second rate. Experiment 1 showed that for both concrete and abstract names, immediate recall of the list was better under imagery mediation instructions than under control instructions. Experiment 2 studied the same conditions using immediate recognition memory of the list as a retrieval measure for the names, and found, despite ceiling effects, that bizarre imagery instructions facilitated recognition for concrete names. Experiment 3 showed that immediate recall could be improved if subjects were given an image mediator for every face-name pair as opposed to generating their own image mediators. Experiment 4 yielded three important findings: (a) 84% of the variance in the 1-week retention of initially recalled names was explained by the presence of absence of the original mediator during 1-week recall; (b) instructions to form image mediators facilitate recall not because image mediators are more effective than other types of mediators, but because they increase the likelihood that a mediator will be formed; (c) 1-week retention could be enhanced with an increased focus during encoding on the points where the mediation process is most likely to fail. The results of these studies are discussed within the context of mediation model wherein recall can fail at any of four stages.