| Abstract: | The subjects (60 boys) were drawn from the sample of a longitudinal study of social development and represented extremely aggressive, anxious, constructive, and submissive behaviour at the age of 8. They were presented with three question series concerning (1) their responses to aggressive attacks; (2) reactions in frustration situations presented in short stories; and (3) their aggressive initiatives. In each series the type of aggressive behaviour, attacker, victim, and other situational factors were systematically varied. In series 2 the type of response, open-ended or forced-choice, was also varied. The results showed that the most valid way of studying boys' self-observations on their aggressive behaviour was to ask if they attack somebody without a specific reason (series 3). This correlated with contemporaneous overt aggression at the age of 8 and predicted aggressiveness and various characteristics of antisocial aggressive development at the ages of 14 and 19. Self-observations on one's physical aggression were more valid for ratings of overt aggressiveness than on verbal aggression. The open-ended or forced-choice type of response did not affect the validity of aggressive responses. Of the categories of nonaggression, ‘conciliatory responses’ had the highest concurrent and predictive validity for constructiveness and other indicators of strong self-control. |
