Information selection and signal probability in multisource monitoring under the influence of centrally active drugs: Phentermine versus pentobarbital

The present study is concerned with the relationship between drug-induced arousal shifts and sampling [(monitoring)] behaviour in a three-source task with an a priori signal occurrence probability of 0.6, 0.3, and 0.1. The multisource monitoring task and procedure was adopted from Hockey (1973) who reported that an arousing treatment, loud noise, produced increased sampling of the course associated with the high signal probability while sleep loss, presumed to decrease arousal level, resulted in a reduction of sampling on this source. Consistent with these findings it was expected that phentermine 20 mg (a stimulant, with a chemical structure related to that of amphetamines) and pentobarbital 100 mg (a barbiturate) would show the same pattern of results, given their opposite effects on brain arousal. Twenty-four male subjects participated in the study and received the drugs according to a placebo-controlled double-blind, three-way crossover design. In contrast to the results from Hockey's study, neither phentermine nor pentobarbital significantly affected average sampling behaviour on the high probability source. Differential drug effects appeared primarily on the low probability source, in that phentermine significantly reduced sampling on this source and pentobarbital produced the highest number of sampling responses to detect signals on this source. The possible reasons for the discrepancy between Hockey's and the present results are discussed.