An Examination of the Structural Validity of the Maslach Burnout Inventory-Student Survey (MBI-SS) Using the Rasch Measurement Model

Jose Manuel Perez-Marmol, Ted Brown

Research output: Contribution to journalArticleResearchpeer-review

Abstract

BackgroundHealth professional students often experience stress and potential burnout when completing their entry-to-practice education. Therefore, having valid scales such as the Maslach Burnout Inventory-Student Survey (MBI-SS) to detect and monitor burnout amongst health professional students is paramount.PurposeThis study investigated the structural validity of the three individual Maslach Burnout Inventory-Student Survey (MBI-SS) subscales. In particular, the dimensionality, scalability, differential item functioning (DIF) based on gender, rating scale structure, and reliability of the three MBI-SS were examined.MethodA sample of 225 Australian occupational therapy undergraduate students completed it. The Rasch Measurement Model (RMM) was used to analyse the MBI-SS item data from the three MBI-SS subscales: Exhaustion, Cynicism, and Professional / Academic Efficacy.ResultsResults supported the unidimensionality and scalability of the three individual MBI-SS and all items from the three subscales met the RMM fit statistics requirements. None of the subscale items exhibited DIF based on gender and the appropriateness of the MBI-SS rating scales were confirmed. The item and person reliability indices were > 0.80 and > 0.77 respectively for each subscale.Discussion and conclusionhe RMM analysis findings indicate that the three individual MBI-SS subscales exhibited adequate levels of structural validity in relation to dimensionality and DIF when completed by occupational therapy students.
Original languageEnglish
Pages (from-to)259-274
Number of pages16
JournalHealth Professions Education
Volume5
Issue number3
DOIs
Publication statusPublished - 15 Sep 2019

Keywords

  • validity
  • burnout
  • Maslach Burnout Inventory-Student Survey
  • Rasch Measurement Model

Cite this