Construct validity of an operant signal detection task for rats.

Many psychoactive drugs produce simultaneous effects on a variety of psychological processes. Behavioral measures in tasks designed to assess cognitive processes in rodents should be able to characterize and dissociate these multiple influences. The present study evaluated how error measures in a classic two-choice operant spatial signal detection paradigm were affected by procedural manipulations of the motivational state of the rat, stimulus properties, and alterations of the inter-trial interval. The experiments were conducted in a two lever operant chamber in which a cue lamp was mounted over each lever. The rats were trained to respond quickly to a short illumination of one of the cue lamps at one of three durations (100, 300 or 1000 ms), presented in a random order. The procedural manipulations were (1) to allow pre-session water access to the normally water-restricted subjects, (2) to vary the intertrial interval (ITI) between sessions, (3) to reduce the intensity of the discriminative stimuli, and (4) to manipulate the variability of the ITI within a session. Stimulus duration-dependent decreases of detection accuracy were observed following pre-session water access and when the intertrial interval was decreased. A reduction of stimulus intensity resulted in decreased accuracy at all stimulus durations. Varying the ITI within the session produced stimulus duration-independent alterations of detection accuracy but no change in the frequency of errors of omission. These findings show that distinct patterns of performance deficits result from manipulating different components of this task and that errors of omission and commission often co-vary and raise questions about the definitions of vigilance and sustained attention as these constructs apply to signal detection tasks that present spatially distinct stimuli.