Longitudinal measurement invariance of the Satisfaction With Life Scale in adolescence
Gepubliceerd in: Quality of Life Research | Uitgave 10/2019Log in om toegang te krijgen
The main purpose of this research was to examine the longitudinal measurement invariance of the Satisfaction With Life Scale (SWLS) in adolescence.
The sample was composed by 484 adolescents from Spain, 46.7% were males. All participants belonged to six academic levels from Grade 7 to Grade 12, and answered the questionnaires at two different times: at the beginning (Mage1 = 14.95, SD1 = 1.81) and at the end of the school year (Mage2 = 15.61, SD2 = 1.81). The reliability of the scale was obtained through Cronbach’s alpha, Guttman lambda, and MacDonald’s Omega total. The multiple group confirmatory factor analysis (MGCFA) was used to examine the fit of the unifactorial model to data and to test the measurement of longitudinal invariance of the scale across two time points (at the beginning T1, and the end T2, of the academic year), and the time points and groups (gender and age).
The values of the single-factor SWLS structure were T1 (CFI1 = 1.000, TLI1 = .997, RMSEA1 = .080, and SRMR1 = .028), and T2 (CFI2 = .997, TLI2 = .995, RMSEA2 = .032, and SRMR2 = .034). On the other hand, values of the reliability and composite reliability when analyzing both time points together as well as separately were as follows: Cronbach’s alpha = .86, Guttman’s lambda = .84, McDonald’s Omega total = .89. Results confirmed the longitudinal invariance of SWLS. The differences in gender and age were not significant and the small differences across time points showed that the means of the latent factor remained the same over time in both variables.
The present study confirmed the single-factor structure of the SWLS in Spanish adolescents, as well as a good reliability and composite reliability. The full longitudinal measurement invariance was also found and there were negligible differences across time points considering gender and age. If these findings are further replicated, the scale could be used to compare the life satisfaction across two time points considering different age and gender groups.