Does PAULA prime PAUL? Does WAITER prime WAITRESS? Implicit transfer across gender forms of verbal stimuli

Repetition-priming effects were investigated with a speeded verification task as an indirect or implicit test. Subjects were asked to verify whether items were instances of the categories "first name" and "profession." In Experiment 1, instances were repeated in their alternate gender form (e.g., PAUL as PAULA, or WAITRESS as WAITER) and performance was compared to new instances. Facilitation in terms of faster response latencies for old instances (i.e., repetitions in the alternative gender form) was restricted to the category "profession." In Experiment 2, facilitation was found for names studied and tested in the same gender form (e.g., PAUL-PAUL). Instances of professions were repeated in either the same or the alternative gender form; the amount of facilitation was identical for the two repetition conditions. However, transfer across gender forms was affected by the study-test order (facilitation was larger from male to female than from female to male) and by the gender dominance of specific instances. In Experiment 3, the interfering effect of gender dominance was not replicated. Response accuracy tended to parallel the latency results in all experiments, but without reaching statistical reliability. Therefore, the faster response latencies for old items are by no means the result of a speed-accuracy trade-off. The central finding of the experiments reported is the material-dependent dissociation of repetition transfer across gender forms that occurred for instances of professions, but was absent for names. This dissociation is interpreted to indicate that the stimulus-driven speeded verification test is sensitive to conceptual, but not to perceptual, processing aspects of a study event.