7. Results
7.1. Data analysis
Hierarchical multiple regression was used to examine the contribution
of age, grit, and community cohesion in predicting prosocial behavior.
Gender, residential location, and years in communitywere included
in the model as control variables because previous research has shown
gender differences in frequency of prosocial behavior (e.g. Caprara &
Steca, 2007), more prosocial involvement by rural residents than
urban ones (Levine, Reysen, & Ganz, 2008), and that sense of community
cohesion and closeness to the community could be affected by years
in the community (Wilkinson, 2008), which could in turn affect frequency
of prosocial behavior. Aiken and West (1991) indicate that in
order to assess interactions, variables must be centered. Therefore, grit
and community cohesion were mean centered before being added to
the equation. In the final step, interaction terms for age with grit and
sense of community cohesion were added to examine age differences
in these variables' prediction of prosocial behavior