All ETDs from UAB

Advisory Committee Chair

Diane M Grimley

Advisory Committee Members

John M Bolland

H Russell Foushee

Joseph E Schumacher

Stuart L Usdan

Document Type

Dissertation

Date of Award

2010

Degree Name by School

Doctor of Philosophy (PhD) School of Public Health

Abstract

A measurement study was conducted to evaluate 11 items representing the latent variable labeled "sense of community" (SOC) employed by investigators of the Mobile Youth Survey (MYS). In previous research the 11 items representing the concept of SOC have been combined to form a composite variable, assuming that the concept was best represented by a single factor (e.g., Bolland, Bryant, Lian, McCallum, Vazsonyi & Barth, 2007). However, no systematic analysis has been conducted to determine the psychometric properties of the 11-item composite variable. The objects of the present study are a) to determine the component structure of the items used to assess SOC among participants of the MYS study across nine years of data collection using principal component analysis procedures; b) to determine the inter-item reliability of the final component solution(s); and c) to establish preliminary validity of the component(s) that emerge by examining convergent and discriminant relationships between the component(s) and MYS psychosocial concepts and behavioral variables. Principal component analysis of data with the total sample demonstrated that a well-saturated, seven-item, two-component structure underlies the 11 items. One three-item component, labeled Individuality, represents respondent perception and judgment of the neighborhood and its social environment. These items demonstrated low levels of internal consistency however, and were not tested further. The other, four-item component, labeled Mutuality, represents friendships and relationships with others in the neighborhood and demonstrated acceptable reliability and validity. Additional analyses demonstrated that the two-component structure remained stable with all subsamples (i.e., both genders and five age categories). Construct validity of the Mutuality scale was supported by mostly significant correlations with selected MYS psychosocial and mental health variables in the predicted directions with the exception of violent behaviors, which may reflect the need for re-examining the role of violence among adolescents in such socially disorganized environments more than raise concerns about the validity of the Mutuality scale. These data support the use of the Mutuality scale as a valid and reliable measure of SOC. Research that employs this measure is needed to determine its utility in studies of SOC.

Included in

Public Health Commons

Share

COinS