Factor structure and psychometric properties of the Pittsburgh Sleep Quality Index in community-based adolescents.

ABSTRACT

Study Objectives: The Pittsburgh Sleep Quality Index (PSQI) is a widely used self-report questionnaire that assesses general sleep quality. This study aimed to validate the single-factor scoring structure and related psychometric properties in the English language version of the PSQI in community-based adolescents. Methods: Participants were 889 (352 males, 39.6%) students (age M = 15.71 ± 1.57; 12.08-18.92 years) recruited from 14 Australian secondary schools. Participants completed the PSQI, Center for Epidemiological Studies-Depression (CES-D) scale, and Spence Children's Anxiety Scale (SCAS). Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) of PSQI component scores were performed on two independent random half-samples (i.e. cross-validation approach). The internal consistency of PSQI components and convergent validity of the PSQI global score with CES-D and SCAS total scores were also assessed. Results: EFA yielded a single-factor model. CFA of the single-factor model in a separate sample yielded acceptable model fit to the data after important relationships were modeled. Namely, modification indices suggested improved model fit by correlating residual scores of PSQI components of sleep duration and sleep efficiency, and sleep efficiency and sleep latency. Internal consistency was acceptable (Cronbach's α = 0.73). The PSQI global score had moderate-to-large positive correlations with CES-D (r = 0.58) and SCAS (r = 0.45) total scores, demonstrating good convergent validity with emotional problems as predicted. Conclusions: The findings validate the single-factor scoring structure of the PSQI in an adolescent sample and highlight important covariation between poor sleep duration, efficiency, and latency in this age group. Further validation studies are required to determine an appropriate PSQI clinical cut-off score for adolescents.