A - 45 Rasch Analysis of the Brief Resilience Scale in Adolescents with Sports-Related Concussion

G D Sanders,L S Hynan,S C Bunt,C M Cullum
DOI: https://doi.org/10.1093/arclin/acae052.45
2024-07-06
Archives of Clinical Neuropsychology
Abstract:Abstract Purpose Accurate measurement of latent abilities is essential for valid inference. The Brief Resilience Scale (BRS) assesses the “ability to bounce back”from stressful situations, and resilience may play a role in concussion recovery. Questions regarding the factor structure of the BRS have been explored in other populations, but evidence of its measurement quality in concussion is lacking. The present study investigated the psychometric properties of the BRS in concussed adolescents within a Rasch framework. Method 927 participants aged 12 to 18 (M = 14.7; 50.3% female) underwent clinical evaluation within 30 days of injury in the ConTex study. The BRS contains 6 items rated on a 5-point Likert scale (Strongly Disagree to Strongly Agree), half of which are reverse scored. The Rasch Rating Scale Model was applied and item fit statistics and dimensionality were examined to assess the degree of fit of the data to the Rasch model. Results Item difficulties ranged −0.82 to 0.77. Item fit was adequate based on mean-square infit and outfit values. The BRS demonstrated sufficient targeting except for participants at the highest and lowest range of latent ability. Unidimensionality testing indicated the presence of a secondary dimension. Conclusions The BRS showed suboptimal fit with the Rasch model based on lack of unidimensionality. However, violations likely reflect item wording effects and not a substantive dimension that merits scale modification. Application of Rasch and other item response theory methodologies can enable development of more psychometrically sound and clinically informative tools for use in the context of concussion outcomes.
psychology, clinical
What problem does this paper attempt to address?