Student's academic efficacy or Inefficacy? an example on how to evaluate the psychometric properties of a measuring instrument and evaluate the effects of item wording

• Autor(a) principal: Maroco, João
• Data de Publicação: 2014
• Outros Autores: Maroco, Ana Lúcia, Campos, Juliana Alvares Duarte Bonini [UNESP]
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• DOI: 10.4236/ojs.2014.46046
• Texto Completo: http://www.scirp.org/Journal/PaperInformation.aspx?PaperID=49248#.Vb-sMBtVhHw; http://hdl.handle.net/11449/126021
• topic: Burnout; Maslach burnout inventory—student survey; Validity; Reliability; Item inversion

This study evaluated the effect of item inversion on the construct validity and reliability of psychometric scales and proposed a theoretical framework for the evaluation of the psychometric properties of data gathered with psychometric instruments. To this propose, we used the Maslach Burnout Inventory, which is the most used psychometric inventory to measure burnout in different professional context (Students, Teachers, Police, Doctors, Nurses, etc…). The version of the MBI used was the MBI-Student Survey (MBI-SS). This inventory is composed of three key dimensions: Exhaustion, Cynicism and Professional Efficacy. The two first dimensions—which have positive formulated items—are moderate to strong positive correlated, and show moderate to strong negative correlations with the 3rd dimension—which has negative formulated items. We tested the hypothesis that, in college students, formulating the 3rd dimension of burnout as Inefficacy (reverting the negatively worded items in the Efficacy dimension) improves the correlation of the 3rd dimension with the other two dimensions, improves its internal consistency, and the overall MBI-SS’ construct validity and reliability. Confirmatory factor analysis results, estimated by Maximum Likelihood, revealed adequate factorial fit for both forms of the MBI-SS (with Efficacy) vs. the MBI-SSi (with Inefficacy). Also both forms showed adequate convergent and discriminant related validity. However, reliability and convergent validity were higher for the MBI-SSi. There were also stronger (positive) correlations between the 3 factors in MBI-SSi than the ones observed in MBI-SS. Results show that positively rewording of the 3rd dimension of the MBI-SS improves its validity and reliability. We therefore propose that the 3rd dimension of the MBI-SS should be named Professional Inefficacy and its items should be positively worded.