AN EVALUATION OF A TRAINING PACKAGE FOR COMMUNITY CANVASSING BEHAVIORS1

Abstract

Basic to the development of most community‐based volunteer programs, neighborhood service centers, and other social service agencies is a program of outreach to the people of the community to be served. Such outreach programs often disseminate information on the service agency by canvassing the residents of the target community. The effectiveness of such programs is sometimes reduced by the failure of the project director to specify appropriate behaviors and the inability of community workers to generate the desired behaviors following only general, unspecific instructions. This study evaluated the effects of an instructional package on the community canvassing behaviors with four trainees: three university field‐work students and a low‐income staff member of a neighborhood service center. The package, consisting of written instructions, quizzes, and feedback, was designed to teach specified target behaviors for presenting information specific to the neighborhood service center (Information Presentation), wearing appropriate apparel (Appearance), making introductory comments (Greeting Behaviors), indicating “interest” in the neighborhood resident (Contact Behaviors), and handling sensitive comments relating to race, religion, and politics (Sensitivity Behaviors). Target behaviors were observed in simulated situations in which scripted statements by a confederate (person playing the role of the neighborhood resident) permitted direct observation of trainee performance in response to a consistent set of conversational stimuli. The effect of the package on the performance of community canvassing behaviors was analyzed by a multiple‐baseline design across behavioral categories. The overall mean percentage of target behaviors increased from 38% during baseline to 94% after training for Information Presentation; from 16% to 97% for Contact Behaviors; and, from 24% to 96% for Sensitivity Behaviors. (The Appearance and Greeting Behaviors of all trainees were at near‐mastery levels before training.) The results suggested that the instructional package was responsible for the increased percentage of target behaviors observed in simulated situations. In addition, the results of two generalization tests suggested that the observed effects were not limited to either a few conversational stimuli or to a single person playing the role of the neighborhood resident. It was concluded that a more comprehensive validation of this social‐skills training program might formally address a number of additional questions: (1) generalization of target behaviors to actual home settings; (2) social validation of the target responses selected for training; (3) generalization to measures of neighborhood resident behavior, and (4) satisfaction of neighborhood residents with the performance of trained community canvassers. This study is suggestive of a technology for systematically validating programs designed to teach complex social skills.