In two-choice tasks, repeating the previous response is beneficial when the task repeats but leads to performance costs when the task switches. Competing accounts explain such response-repetition costs in different ways. Among these, the signalling account posits that changing the task “signals” the cognitive system to change the response as well. With two response alternatives, a task switch correctly signals the alternative response, yielding response repetition costs when the response should instead be repeated. However, with three response alternatives, a task switch favours a general response switch bias, which does not prime a specific alternative response, hence it should not yield the same response repetition costs. We therefore manipulated the number of response alternatives in a cued task-switching paradigm, with one session affording two response alternatives and another affording three. This manipulation did not strongly influence reaction time measures of response-repetition costs. It did, however, influence response-repetition costs in mean error rate, which were smaller with three – relative to two – response alternatives. The latter finding provides support to the signalling account over the alternative accounts, since this reduction is predicted uniquely by the signalling account. We discuss this finding in light of other empirical results suggesting similar mechanisms, namely promoting a transition-matching bias, wherein performance improves when different elements of a task require the same transition (e.g., a repetition), compared to a different transition.

The zipped project folder contains one README file (README.pdf), one data file with the clean data (DEI11_Pro_all_pps11-12.csv),one R script with the analyses (data_cleaning+analyses.R) and one subfolder with the raw data (raw_data).