The potential of correspondence training for facilitating generalisation of social skills.

Recent research suggests that the effects of social skills training programmes are often not evident outside the training setting. Correspondence training has been advocated as a means of overcoming this problem, although few studies have demonstrated its effectiveness. Three intellectually handicapped adults were given social skills training. At other times, generalisation was tested during conversations with staff. Subsequent to mastery in the social skills training setting, behaviours that did not occur at criterion in the generalisation setting were selected for correspondence training. After correspondence training was introduced, targetted behaviours increased in frequency in the generalisation setting. These increases were maintained during follow-up sessions, which took place 1 month after the training programme ceased. The effects of correspondence training are examined in the context of previous reinforcement history, accuracy of reporting, and verification of performance in unobserved settings.