Big data analytics quality factors in enhancing healthcare organizational performance: a pilot study with Rasch model analysis

Big Data Analytics (BDA) plays a pivotal role in the digital transformation of healthcare, significantly boosting organizational performance within the sector. As healthcare organizations increasingly adopt BDA to leverage data-driven decision-making, understanding the factors contributing to BDA quality becomes imperative. Thus, this study has proposed and developed the BDA quality conceptual model, and a pilot study is part of the process of completing the conceptual model development. The instrument, which is the questionnaire that has been designed, needs to be tested for reliability. Therefore, the pilot study aims to evaluate and refine the instrument used to assess BDA practitioners’ comprehension of the constructs and the reliability of the items. This study utilized the probabilistic approach of Item Response Theory (IRT), explicitly employing the Rasch Measurement Model analysis to enhance the accuracy of measurement instruments, assess respondents' performance, and ensure instrument reliability. The survey instrument comprised 11 constructs and 64 items, which were designed to measure all the constructs: reliability, accuracy, completeness, timeliness, format, accessibility, usability, maintainability, portability, user satisfaction, and healthcare organizational performance. Data were collected from 20 respondents and synthesized according to their responses to each questionnaire item. The analyses were performed using Rasch analysis software, specifically Winsteps. The results of the Rasch analysis included findings on the reliability of persons and items, the distribution map of person-item relationships, identification of misfitting items, and assessment of unidimensionality. Ten items were removed from the initial set of 64 due to misfit, leaving 54 items that effectively measured respondents' understanding of BDA quality in healthcare organizational performance. Thus, Rasch measurement model analysis has confirmed the instrument was well constructed, valid, and reliable for actual study.