A Longitudinal Investigation of the Satisfaction With Life Scale Among Adults with Severe and Persistent Mental Illness Participating in Psychosocial Rehabilitation Services

Background and Purpose: Life satisfaction has been recognized as an important construct for understanding the process and outcome of recovery among individuals with severe and persistent mental illness (SPMI). An increasing number of intervention studies have used life satisfaction as an outcome measure. Yet, little is known about the psychometric properties of the Satisfaction With Life scale (SWL; Stein & Test, 1980) when used with people with SPMI. In order to use life satisfaction as a useful and valid outcome measure to assess change following an intervention, it is critical that the instrument measures the same construct at different occasions. If this stability of a construct, or “measurement invariance” does not hold over time, observed changes may be confounded by the instability of the construct. The purpose of this study is to examine the measurement invariance of the Satisfaction With Life scale in order to validate its use in the context of longitudinal mental health intervention research in adults with SPMI participating in community-based psychosocial rehabilitation services.

Method: The sample consisted of 331 individuals with schizophrenia spectrum disorders who participated in psychosocial rehabilitation services in Los Angeles, California. The sample was followed prospectively for twelve months, with psychosocial observations made at baseline and every six months. The Satisfaction With Life is a 18-item self-report instrument that measures subjective aspects of quality of life in four domains (living situation, work, socialization, and self/present life). Using the original four-factor model (Test et al., 2005) as a baseline model, a series of multiple-group confirmatory factor analyses (CFA) were conducted to test five types of measurement invariance (configural invariance, weak factorial invariance, strong factorial invariance, factor covariance/error invariance, and strict factorial invariance) on three waves of data. A Chi-square difference test was used to examine the equivalence of parameters in nested measurement models across time points.

Results: Results of CFA supported the originally hypothesized four-factor model of the SWL (c˛/df =891.302/387, CFI=.916, TLI=.900, RMSEA=.068). Therefore, the model was selected as the baseline model for subsequent measurement invariance tests. Results of multiple-group CFA revealed that all types of measurement invariance held over time. The most restrictive type of invariance (strict factorial invariance) was achieved (c˛/df = 1025.265/499, CFI=.912, TLI=.919, RMSEA=.061).

Conclusions and Implications: This study demonstrates that the most stringent type of invariance holds across time. This finding indicates strong psychometric properties of the SWL scale when used with adults with SPMI. The SWL scale appears to be a promising instrument to assess change in life satisfaction in longitudinal studies. The demonstration of longitudinal invariance is an important first step in examining the change in life satisfaction longitudinally among consumers who receive community mental health services. Beyond the usefulness of the SWL in intervention research, the SWL can be used as a brief assessment or screening too in clinical settings.