Relation between stimulus discriminability and response bias in a delayed conditional discrimination task in the rat

The present study investigated whether changes in response bias (log b) affect discriminability (log d) in a delayed conditional discrimination (DCD) task. Eight rats were trained on an DCD task and response bias was manipulated by changing the reinforcement ratio (RR) for correct responding on the two response alternatives. Three findings emerged. First, changes in RR produced changes in log b and the magnitude of these changes was related to the value of the RR. Second, log d was affected by changing RRs. Third, the effect of changes in RR on log b was larger after longer delays. These results suggest that log d and log b are not independent and it is concluded that changes in response bias (log b) may affect the discriminability measure log d. This implies that changes in log d may be mediated by changes in response bias and it is therefore recommended that log d and log b values are both reported when investigating effects of, for example, pharmacological manipulations on DCD performance.