Do Missing Domain Scores Compromise the Validity of the Quality of Life Inventory?

Abstract

Background and aims: The Quality of Life Inventory (QOLI, Frisch, 1994) manual states that in most cases QOLI total scores are invalid when two or more of the 16-domain scores are missing. The current study aimed to investigate this guideline.Methods: Two samples were utilised consisting of 259 community-dwelling adults and 144 adults surveyed 12 months following traumatic brain injury (TBI). First, the domains of the QOLI were regressed against Quality of Life Index (QLI) total scores. Second, a series of Receiver Operator Curve analyses systematically investigated the sensitivity of QOLI scores in detecting depression, as identified by the HADS and DASS.Results: The final model predicting QLI scores comprised seven of the 16-QOLI domains, R2 = .57, and accounted for equivalent variance to the full 16-domain model, R2 = .59. With as few as seven domains, the sensitivity of QOLI scores in identifying participants with depression was very good and equivalent to the complete 16-QOLI domain total score (>76%). Similar results were observed when these analyses were replicated within the sample with TBI.Conclusions: These findings showed the QOLI was more robust to missing domain scores than the current validity guidelines stated in the scale's manual suggest. Future research could determine the core domains of the QOLI in a range of samples including adolescents and specific clinical groups.