Paper
Friday, 21 July 2006
This presentation is part of : Developing Measurement Instruments
Exploratory Factor Analysis of the Australian/English Version of the Diabetes Management Self-Efficacy Scale
Jan McDowell, RN, PhD1, Mary Courtney, RN, PhD2, and Helen Edwards, RN, PhD2. (1) Institute of Health and Biomedical Innovation, Queensland University of Technology, Kelvin Grove, Australia, (2) School of Nursing, Queensland University of Technology, Brisbane, Australia
Learning Objective #1: identify an instrument that is available to measure multiple aspects of self-efficacy towards diabetes self-management.
Learning Objective #2: understand the process involved in undertaking factor analysis of data.
The 20 item Australian/English version of the Diabetes Management Self-Efficacy Scale (DMSES) is a valid and reliable instrument tailored to measure self-efficacy towards managing diabetes-related self care within the Australian environment.  To date the tool has been regarded as a unidimensional scale. This paper presents results from exploratory factor analyses undertaken to determine if the instrument measures multiple aspects of self-efficacy in this context.   One hundred and eighty-four community-dwelling Australians with type 2 diabetes (53% female) were recruited to this cross-sectional study via advertisements in magazines, medical practices, and community health centres. Participants completed a mailed self-report questionnaire that included the DMSES. The sample was middle aged (M=63 years; SD = 10.5), retired (81%), and had mostly been diagnosed with diabetes for more than a year (74%). The data met assumptions for factor analysis and the sample size was adequate (KMO = 0.9).  Factors were extracted using Principal Component Analysis plus Varimax with Kaiser Normalisation. With eigenvalues set at greater than 1, the initial analysis identified four factors. However, all but four scale items loaded at 0.4 or above on the first factor, suggesting that the DMSES may be unidimensional. Subsequent analysis, during which the data were rotated, identified four disparate factors with items loading at 0.53 or above. Factor 1, comprising nine items related to ‘food and diet’, explained 26.7% of the scale variance. Factor 2 contained five items associated with ‘doing diabetes-related activities’ e.g. examining feet (17.3% variance). Factor 3 included three items related to ‘checking blood glucose level’ (13.3% variance); and Factor 4 consisted of 3 items associated with ‘taking medication and visiting the doctor’ (10.2% variance). Outcomes indicate that this version of the DMSES contains four subscales that measure self-efficacy towards undertaking aspects of diabetes management. Confirmatory factor analysis is now required to validate these findings.

See more of Developing Measurement Instruments
See more of The 17th International Nursing Research Congress Focusing on Evidence-Based Practice (19-22 July 2006)