This paper analyzes the dataset collected from students participating in the Boy With A Ball (BWAB) program, a faith-based community outreach group, through the Hemingway Measure of Adult Connectedness©, a questionnaire measuring the social connectedness of adolescents. This paper first approaches the data in the conventional method provided by the Hemingway website. Then it identifies which questions are strong determiners in deciding whether a student has completed the BWAB program or not. With the goal of utilizing the logistic regression, the set of questions to those only identified as significant in other methods is reduced. These methods include linear regression, decision tree, and random forest. The results are explained from a psychological perspective of social adolescent development.