Effects of sense of community and other individual and group variables on duration of community involvement: A survival analysis

Sense of community has been characterized as an interactive phenomenon: a dynamic relationship likened to the interplay of figure and ground within a visual field. Studies on sense of community have typically focused on the attributes of specific social contexts. A shortcoming of such research has been the presupposition that participants' sense of community experience is limited to one social context. The effects of individuals' past community experiences, and other psychological variables, have been overlooked as determinants of community involvement.;The current study utilized a formulation of sense of community as a personality variable, and employed increased focus on psychological variables in general. A new sense of community scale was used. This scale was applicable to both territorial and non-territorial community affiliations. The study's first hypothesis predicted that both demographic and psychological variables would be significantly related to length-of-residence at Oxford House; the second hypothesis predicted that both cohort group characteristics and variables representing interactions between individuals and their cohort groups would significantly predict length-of-residence.;Individual interviews with 129 male Oxford House were used to collect data relevant to the set of predictor variables. A questionnaire constructed to standardize the interviews comprised measures of four psychological variables--sense of community, stress, social support, and hope--as well demographic items. Interview data were used to generate survival functions corresponding to the predictor variables, and to complete two Cox Regressions.;Results from the Cox Regression analyses indicated that the age, race, and level of education, as well as cohort group age, difference between individual and cohort age, percentage of African American cohorts, and percentage of White cohorts were all significant predictors of length-of-residence at Oxford House. Kaplan-Meier survival analyses were used to inform the interpretation of these results, and also suggested the novel finding that participants' level of pessimism was negatively related to length-of-residence. In addition, post hoc analyses of participants' sense of community scores gave some support to the notion that sense of community acts, to a degree, as a personality variable, manifesting a level of consistency across social contexts.