John Whitney and Dr. Renata Forste, Sociology Department

This study was designed as a program assessment for school-aged youth attending Boys and Girls Clubs in the greater Salt Lake City area. The study focused on the relationship between length of membership at the club on behavioral outcomes particularly with the goal of a reduction of delinquent behaviors. The original study was designed to be a longitudinal study of club participants whereby the participant’s responses were recorded using their club ID. These IDs would be used to track how individual respondent’s attitudes and behaviors changed over the course of two years as a result of their club participation. Due to problems in the data collection process, however, an insufficient number of respondents were represented in both waves of the data to complete the desired analysis. Some of the possible reasons for the lack of overlapping responses may include inaccurate record keeping on part of the club, errors on the part of club participants in recording their ID numbers, or may be the result of lower frequency of attendance by club members by which respondents from previous waves were absent during the data collection process during this wave. Higher mobility among at risk families may also play a role in the data problems as many club attendees report both high levels of weekly attendance and long duration of attendance in years but do not have data during both data points because they may have attended two different clubs and have been issued different ID numbers because of the move. As a result of these data issues the two waves of data were combined into a single larger sample with duplicate cases removed.

Another modification to the study’s design came as a result of additional research that occurred after the proposal was submitted. One key study indicated that the relationship between attendees and staff may account for many of the beneficial results observed from clubs and afterschool programs (Pierce, Bolt, & Vandell 2010)1. As a result, a series of questions asked during both waves of the study were incorporated into the analysis as a new key predictor variable.

Results from using this new variable were somewhat consistent with prior research. A club member’s attachment to club staff was significantly related to reduced delinquent behavior which consisted of substance abuse (alcohol, drug use, and smoking) and incidence of theft in the last thirty days. However, staff attachment is not significantly related to reduced delinquent behavior once variables measuring club attendee’s moral beliefs were added to the model. Other variables were added to test their significance such as gender, race, length of membership, age, and the club to which the member belongs. Of these, club of attendance was the only variable that was significantly related to delinquent behavior so the other variables were removed from the analysis. The significant results from which club the youth were enrolled at showed that members of clubs that were closer to downtown Salt Lake City were significantly more likely to engage in delinquent behaviors than clubs that were further away from the city. Since three of the four variables used to measure delinquency were related to substance usage it is possible this results from availability of substances and pressure from peer groups.

One of the conclusions drawn from this analysis is that the connection between club attendees connect and staff personnel appears to be a more accurate indicator of improvement than even length of membership when measuring reductions in delinquent behaviors. This contrasts with the original hypothesis that length of membership in the club was the best measurement for improvement.

To be certain that these results are genuine, however, improvements need to be made to the data collection process that allows for respondents to take the survey over the course of several weeks to include respondents from prior waves of the study. The club member tracking system could also be improved in order to ensure that the same number is not issued multiple times and that respondents have only one number even if they attended different clubs, especially when membership in multiple clubs could be an issue. Having longitudinal data of this time will make it possible to determine if respondents are experiencing genuine growth or if the effects observed are the result of the good members (youth engaging in fewer delinquent behaviors) remaining in the program while the at risk members are dropping out.

The research model for this study is still being tested. A more detailed report of findings is being drafted and will be submitted for publication later in the summer.

References

 

• Pierce, Kim M., Daniel M. Bolt, Deborah Lowe Vandell. 2010. Specific Features of After-School Program Quality: Associations with Children’s Functioning in Middle Childhood