Three reliably measured components of conversation-questioning, providing positive feedback, and proportion of time spent talking-were identified and validated as to their social importance. The social validity of the three conversational behaviors was established with five female university students and five female junior-high students. Each was videotaped in conversations with previously unknown adults. The conversational ability of each girl was evaluated by a group of 13 adult judges who viewed each tape and rated each conversant "poor" to "excellent" on a seven-point rating scale. The average ratings of the girls correlated at r = 0.85 with the specified behavioral measures. These procedures were replicated with additional subjects and judges and yielded a correlation of r = 0.84. The high correlations between ratings and the objective measures suggested that the specified conversational behaviors were socially important aspects of conversational ability. Employing a multiple-baseline design across the behaviors of asking questions and providing positive feedback, an attempt was made to train four girls who used these behaviors minimally to engage in the behaviors in conversations with adults. Adult judges were again employed to rate randomly selected samples of the girls' skills in pre-and posttraining conversations. The average ratings of the girls before training were lower than both the university girls and the junior high-school girls. After training, the girls' conversational abilities were rated substantially higher than those of their junior high-school peers. These rating data validated the benefits of the training and the social importance of the behavioral components of questions and feedback in conversation. The authors suggest that it may be necessary for traditional behavior analysis measurement systems to be supplemented by social-validation procedures in order to establish the relationship between "objectively" measured behaviors and complex classes of behavior of interest to society.

A survey of Lost and Found classified sections in metropolitan and smaller newspapers revealed disparate rates between Lost ads and Found ads: Lost ads greatly outnumbered Found ads, probably because newspapers usually require the finders of lost personal property to pay for Found advertisements. The effect of a Free-Found-Ad policy on the rate of Found advertisements placed in the Lost and Found sections of three community newspapers was investigated using a multiple-baseline design. The results suggested that the Free-Found-Ad policy was effective in increasing the rates of Found ads in all three newspapers. To determine whether increases in Found ads resulted in increases in recovered property, a sample of individuals who placed Found ads were surveyed in both baseline and treatment conditions and asked if the found items had been claimed by their owners. The Free-Found-Ad policy appeared to be effective in increasing the amount of personal property returned. The study concluded that community newspapers can provide incentives to increase such help-giving or altruistic behaviors. The implications of this study for a general policy-research strategy are discussed.

Three experiments were conducted (1) to identify the important behaviors of police-youth interaction, (2) to teach four of these behaviors to three court-adjudicated delinquent youths, and (3) to validate the changes through the use of subjective ratings by police officers and a group of citizens from the community. The data suggest that training improved the police officers' subjective impressions of the youths and may have decreased the probability of the youths' being taken into custody, being stopped again, or being labeled as troublemakers.