Community volunteers as agents of rehabilitation

This investigation used a pre-test, post-test; two-group, randomized design to test the effect of the assignment of a "Big Brother" on juvenile probationer’s self-esteem scores. Results were not statistically significant, indicating that the volunteer manipulation did not affect self-esteem scores. Ex post facto power tests suggested that future studies will need a larger number of subjects, a more homogeneous population, if possible, a more sensitive measure of self-esteem and an increase in the strength of the manipulation. The manipulation can be strengthened in future studies by a longer time span and tighter control over volunteer activities. In many cases volunteers saw subjects only once or twice and in some cases not at all. It is suggested that this can be controlled by training and supervision of volunteers. Volunteers were also tested using a self-concept measure and an internal-external control measure. No significant correlation was found between volunteers’ scores and juveniles' self-esteem change scores. Two important results were: a new awareness of the need for supervising volunteers and the proof that a tightly controlled randomized design can be implemented in a community setting, with juvenile delinquents.