The BETA® nursing measure: Calibrating construct validity with Rasch analyses

BackgroundThe BETA nursing measure has been introduced as a tool to routinely measure and monitor the outcomes of patients' activities of daily living in a restorative nursing care context.ObjectivesTo investigate the BETA's construct validity using the Rasch model with specific reference to the BETA's potential to be used as an interval scale providing metric or interval data.MethodA quantitative analytical design was followed using Rasch analyses whereby BETA raw data was collected from patients (n = 4235) receiving nursing care in 28 South African sub-acute and non-acute nursing facilities. The data was prepared for Rasch analyses and imported into WINSTEP® Software version 3.70.1.1 (2010). Final results were shown by means of figures and graphs.ResultsA successful outcome was achieved by dividing the BETA into four subscales. In this process one of the original BETA items was omitted and seven other items required collapsing of their categories before the four subscales achieved a satisfactory fit to the Rasch model.ConclusionThe four BETA subscales achieved “very well” to “excellent” levels of fit to the Rasch model. This finding thus creates an opportunity to convert the BETA's Likert qualities into an interval measure to calculate change in patients' activities of daily living metrically as a direct result of effective restorative nursing.