Measuring sense of community: A methodological interpretation of the factor structure debate

Instability in the factor structure of the Sense of Community Index (SCI) was tested as a methodological artifact. Confirmatory factor analyses, tested with two data sets, supported neither the proposed one‐factor nor the four‐factor (needs fulfillment, group membership, influence, and emotional connection) SCI. Results demonstrated that the SCI was best described by four factors in combination with two method factors (i.e., one factor including only positively worded items and another including only negatively worded items). Analyses comparing SCI and method factors showed that positively and negatively worded items correlated differently with measures of citizen participation, psychological empowerment, and health‐related quality of life. Results indicate that the SCI should be revised to exclude negatively worded items and new positively worded items should be developed and tested. The effects of method variance and alternative solutions to the psychometric issues in the SCI are discussed. © 2006 Wiley Periodicals, Inc.