Context-specific control in the single-prime negative-priming procedure

Abstract

The current paper examines the applicability of the context-specific control principle to the probe selection dependence of negative-priming effects using the single-prime procedure. In a series of experiments, we highlight the applicability of the context-specific control principle, first by illustrating a key result that implicates the role of context-specific control and challenges the contextual similarity principle. Following this, we show the importance of distinct probe contexts in the single-prime negative-priming procedure and report a novel finding that illustrates a learning effect that can occur within an experimental session. Finally, we test the relation of our novel learning effect to a related learning proposal offered by Frings and Wentura (2006), and we demonstrate that the learning involved in context-specific control is not dependent on contingency learning. Overall, the patterns of results highlight the role of context-sensitive memory in controlling how current perception and action are integrated with prior experience.

Keywords:

• Episodic memory
• Selective attention
• Learning
• Negative priming
• Selection difficulty

Acknowledgments

This research was supported by a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant awarded to B.M. and by an NSERC CGS doctoral scholarship to M.C.D.

Notes

¹ Given our interest in the effect of learning in the context of priming effects, we analysed the effect of repetition on low-selection probes in the second half of the experimental session. This analysis revealed a significant effect of repetition, F(1, 13) = 4.86, MSE = 2,493.526, p = .046. Participants responded 21 ms faster for repeated than for not-repeated trials; that is, a significant positive-priming effect was measured.

² This F-test is equivalent to a one-sample t test for the 10-ms repetition effect being significantly different from zero. Given our specific a priori prediction, a one-tailed test seems appropriate.