Influence of Mapping Complexity on Negative Priming for Incompatible Spatial Mappings

Article excerpt

For tasks with an incompatible stimulus-response mapping, whether the compatible response must be inhibited is an ongoing issue. Read and Proctor (2004) tested this inhibition hypothesis, using a negative priming paradigm for four-choice tasks with three different incompatible spatial mappings. For a mapping that did not follow a simple rule, reaction time was lengthened when the corresponding response on the preceding trial became the required response on the current trial, as compared with when it did not, showing a negative priming effect. However, for mappings that followed a simple rule, negative priming was not evident. The present study extends this research to a more complex mapping. On the basis of a two-process model adopted from the negative priming literature, we hypothesized that high mapping complexity should also diminish the negative priming effect for incompatible mappings, because the balance of cognitive resources is allocated to identification of the correct response. Two experiments are reported in which mappings of different complexity were used in six-choice spatial tasks. Analyses of reaction times showed that negative priming diminished with increased mapping complexity, apparently due to increased dominance of response identification processes, rather than inhibition of the corresponding response.

Since its inception, negative priming (NP) has been considered a useful tool for understanding the role of inhibition in control of action (Fox, 1995). In an NP task, an irrelevant stimulus or dimension on the previous trial (prime) becomes relevant on the current trial (probe). Reaction time (RT) for these trials is lengthened, as compared with trials on which the prime and probe are unrelated. This lengthening is most often attributed to residual inhibition of the irrelevant information on the prime impeding activation of the correct response to the probe (Tipper, 1985). NP has been demonstrated in many tasks, including object identity, letter identity, and semantic associations (Fox, 1995). However, the increased variety of NP tasks has shown mixed results. Depending on task requirements, NP is modulated and even reversed, and alternative accounts have been proposed that do not require inhibition, instead emphasizing retrieval (e.g., Neill, Valdes, Terry, & Gorfein, 1992).

Recently, NP has been demonstrated in stimulus-response compatibility (SRC) tasks. SRC refers to the finding that responses are faster when stimuli and their assigned responses correspond than when they do not. This phenomenon has been widely used to examine issues of response activation and selection in numerous task environments, as well as for various participant populations (see Proctor & Vu, 2006). However, as with NP tasks, results in SRC tasks are subject to alternative interpretations (Read & Proctor, 2004; Shiu & Kornblum, 1996).

One major question is whether inhibition of the corresponding response is necessary for response selection when the stimulus-response mapping is incompatible. Incompatible mappings necessitate that participants adopt spatial transformations requiring use of short-term, taskdefined stimulus-response associations that are different from the established long-term associations between stimuli and their corresponding responses. Consequently, participants may need to inhibit the tendency to make the corresponding response activated through the long-term associations (Kornblum, Hasbroucq, & Osman, 1990). An advantage of the NP paradigm for studying SRC effects is that it provides a measure of residual inhibition of the corresponding response. However, as was indicated, NP in both basic NP tasks and SRC tasks can be dominated by the memory requirements of the task (e.g., Chao & Yeh, 2008; Proctor, Vu, & Marble, 2003).

A Two-Process Model for SRC

Two-process models have been proposed for both SRC and NP. In the compatibility literature, Kornblum et al. …