Assessing Social Connectedness Among Persons with Schizophrenia: Psychometric Evaluation of the Perceived Social Connectedness Scale

Background/Purpose:

Social connectedness is defined as the perception that one is part of a mutually supportive social system. Few studies have examined socially supportive networks in adults with schizophrenia and co-morbid chronic health conditions (e.g., hypertension) receiving integrated health services in community-based settings. The purpose of the current study was to evaluate the psychometric properties of the perceived social connectedness (PSC) scale by examining its factor structure, reliability, and construct validity. Anchored in social capital theory, interrelationships among the PSC scale and theoretically linked measures of symptom severity and perceived daily functioning were examined to test convergent validity. The relationship between the PSC scale and a theoretically distinct variable, type of health insurance, was observed to assess discriminant validity.

Method:

A total of 146 adults diagnosed with schizophrenia comprised the analytic sample. Measures included the four-item PSC scale, the eight-item perception of functioning (PF) scale, and the six-item K6 for symptom severity. Randomly split subsamples were used to conduct the evaluation. Exploratory factor analysis (EFA) was employed with subsample 1 (n₁=73), whereas reliability and construct validity were assessed with subsample 2 (n₂=73). Prior to conducting EFA, the Kaiser-Meyer-Olkin (KMO) and Bartlett’s sphericity tests were computed to ensure sample suitability. Inter-item consistency of the PSC scale was assessed with Spearman’s rho and overall reliability was examined via Cronbach’s alpha and Guttman split-half coefficients. Construct validity of the scale was assessed via expected convergence with symptom severity and daily functioning and absence of convergence with type of health insurance.

Results:

Participants were approximately 47 years old. The majority identified as male (59%) and African American (78%). The KMO and Bartlett’s sphericity tests showed strong results for sample suitability, at 0.777 and p<0.001, respectively. The principal axis factoring on the PSC scale revealed a single dimension with an eigenvalue over 2.6. The unidimensional solution accounted for 54.2% of the common variance, whereas the single factor explained 65.3% of the total variance. All four items (friendships, support for enjoyment, belonging, support during crises) loaded soundly on the lone factor, at 0.764, 0.713, 0.825, and 0.631, respectively. PSC scale inter-item correlations ranged from 0.515 to 0.617, at p<0.001. A Cronbach’s alpha of 0.790 was observed for overall reliability with similar results obtained with Guttman split-half coefficients: λ-4=0.777, with the highest of the lower-bound lambdas, λ-2, at 0.791. Results of validity testing of the PSC scale showed evidence of convergent validity with both the PF and K6 scales (rs=0.4666 and -0.393, respectively, ps<0.001). Discriminant validity of the PSC scale was indicated by its absence of association with type of health insurance (r=0.072, p=0.556).

Conclusions/Implications:

With findings of factor structure, reliability, and validity, the current study demonstrates that the PSC scale is a useful tool for assessing individual perceptions of social support in persons with schizophrenia receiving integrated care. The four-item PSC scale is brief and shows preliminary evidence of strong reliability and validity. The PSC tool provides health professionals with information that can guide efforts to promote social connectedness and inclusion among at-risk clients with schizophrenia.