Rasch analysis of the postconcussive symptom questionnaire: measuring the core construct of brain injury...

The Postconcussive Symptom Questionnaire (PCSQ; Lees-Haley, 1992 ) is purported to measure four constructs. These include psychological, cognitive, somatic, and infrequency (i.e., items intended to reflect negative impression management) symptoms. The utility and validity of Postconcussive Syndrome (PCS) as a diagnostic condition continues to be debated. To this end, examining the instruments used to measure postconcussive symptoms can increase our understanding with respect to this issue. The aim of this study was to derive a revised PCSQ to target the core construct of subjective symptoms reported by persons with traumatic brain injury (TBI). A total of 133 people with mild to severe TBI completed the 45-item PCSQ. Items were scored dichotomously, as symptom present or absent. Rasch analysis, based on the mathematical model formulated by Rasch ( 1960 ), was used to derive the revised PCSQ. Misfitting and redundant items were removed and a second model containing 19 items was fitted. The revised PCSQ-19 had superior psychometric qualities; reliability was 0.81. The PCSQ-19 provides a more targeted, unidimensional assessment of subjective symptoms following brain injury. The findings also revealed information related to symptom hierarchy which can further our understanding of PCS.