Holger Muehlan1, Lindsay Wilson2, Nicole von Steinbüchel3. 1. Ernst-Moritz-Arndt University of Greifswald, Greifswald, Germany holger.muehlan@uni-greifswald.de. 2. University of Stirling, Stirling, UK. 3. University Medicine of Goettingen, Goettingen, Germany.
Abstract
OBJECTIVE: The Quality Of Life after BRain Injury (QOLIBRI) consortium has developed a short six-item scale (QOLIBRI-OS) to screen health-related quality of life after traumatic brain injury. The goal of the current study is to examine further psychometric qualities of the Quality Of Life after BRain Injury-Overall Scale (QOLIBRI-OS) at the item level using Rasch analysis with particular emphasis on the operating characteristics of the items. METHOD: A total of 921 participants with traumatic brain injury were recruited. The analysis sample was restricted to 795 participants with Glasgow Coma Score and Glasgow Outcome Score-Extended available in order to ensure a well-characterized sample. RESULTS: Overall fit statistics indicate sufficient reliability of the QOLIBRI-OS. The assumption of unidimensionality could be confirmed with reservation. The range of item locations is small, whereas item thresholds cover a wide range of the latent trait. The majority of parameter estimations for all class intervals of the respective test are in accordance with the model assumptions. CONCLUSION: The results show that, despite marginal misfits to the model, the six items representing the QOLIBRI-OS could establish a Rasch scale.
OBJECTIVE: The Quality Of Life after BRain Injury (QOLIBRI) consortium has developed a short six-item scale (QOLIBRI-OS) to screen health-related quality of life after traumatic brain injury. The goal of the current study is to examine further psychometric qualities of the Quality Of Life after BRain Injury-Overall Scale (QOLIBRI-OS) at the item level using Rasch analysis with particular emphasis on the operating characteristics of the items. METHOD: A total of 921 participants with traumatic brain injury were recruited. The analysis sample was restricted to 795 participants with Glasgow Coma Score and Glasgow Outcome Score-Extended available in order to ensure a well-characterized sample. RESULTS: Overall fit statistics indicate sufficient reliability of the QOLIBRI-OS. The assumption of unidimensionality could be confirmed with reservation. The range of item locations is small, whereas item thresholds cover a wide range of the latent trait. The majority of parameter estimations for all class intervals of the respective test are in accordance with the model assumptions. CONCLUSION: The results show that, despite marginal misfits to the model, the six items representing the QOLIBRI-OS could establish a Rasch scale.
Authors: Cindy Hunt; Shatabdy Zahid; Naomi Ennis; Alicja Michalak; Cheryl Masanic; Chantal Vaidyanath; Shree Bhalerao; Michael D Cusimano; Andrew Baker Journal: Qual Life Res Date: 2019-09-14 Impact factor: 4.147