Reliability across studies from the functional assessment of cancer therapy-general (FACT-G) and its subscales: a reliability generalization

Abstract

Purpose: The aim of this study was to conduct a reliability generalization of the Functional Assessment of Cancer Therapy-General (FACT-G) and its subscales to examine variation in score reliability across all published studies.

Methods: We reviewed 344 publications based on predetermined criteria. About 78 published studies reported Cronbach's Alpha reliability coefficients from their study in which data were collected. Sample size based weights were applied, and studies were coded on several scale and demographic characteristics. Using independent samples t tests, we examined associations between study characteristics and internal consistency variability.

Results: Average FACT-G score reliability was .88 (subscales ranged between .71-.83). Three variables produced small, statistically significant (P < or = .05) eta squared effects (ranging between .06-.21) due to different sources of variation in the FACT-G and subscales: ethnicity, cancer type, and study type-all of which appeared to be related to disproportionate representation of studies with the majority including Caucasian samples, mixed cancer samples, and validation type studies.

Conclusions: The FACT-G and its subscales demonstrated acceptable reliability evidence across observed studies, without substantial variability due to scale or demographic characteristics.