Preference for predicted over unpredicted shock

Rats received unavoidable, inescapable shock at each end of a shuttle box. A predictive signal preceded the shock when the rat was at one end of the shuttle box, while a random signal was presented when the rat was at the other end. Subjects spent significantly more than 50% of their time at the end with the predictive signal. Examination of crossover data suggested that the occurrence of crossover responses from one end to the other was influenced by the consequences of those responses, and raised the possibility that preference for predicted shock in this and previous studies was due to differential consequences that normally exist for crossover responses in each direction.