Paper
Friday, 21 July 2006
This presentation is part of : Measurement: Design and Development of Instrumentation to Assess Patient Attitudes and Beliefs and to Collect Background Information
Development and Testing of an Appropriate Behavior-Specific Self Efficacy Instrument
Judith A. Erlen, PhD, RN, FAAN, Health Promotion and Development, University of Pittsburgh, Pittsburgh, PA, USA, Donna Caruthers, NA, University of Pittsburgh, Pittsburgh, PA, USA, and Lora Burke, PhD, RN, Health and Community Systems, University of Pittsburgh, Pittsburgh, PA, USA.
Measuring self-efficacy, an individual's perception of how capable he or she is to perform a specific behavior (Bandura), requires behavior-specific instruments. The purposes of this paper are to (1) describe constructing and scaling of items for three self-efficacy tools, and (2) examine the psychometric properties of three investigator-developed self-efficacy instruments. Each instrument was grounded in the literature and used an expert panel for content validity. The first scale focuses on medication taking behavior in patients with HIV infection (HIV Medication Taking Self-Efficacy Scale). The sample included 215 subjects with HIV (67.4% male; 54% white: average age 40.66±5.58 years). Cronbach's alpha was .96 and test-retest over 612 weeks was r=.58. The second scale focuses on tobacco abstinence among hospitalized patients (Tobacco Abstinence Self-Efficacy Scale). The sample included 80 hospitalized smokers (40% male; 79% white; average age 51±11.18 years). Cronbach's alpha was .95; test-retest over 12 weeks was r=.73. The third scale focuses on adherence to a cholesterol-lowering diet (Cholesterol-Lowering Diet Self-Efficacy Scale-Short Form). The sample included 238 (51% male; 94% white; average age 56±11.5 years) adults with hyperlipidemia. Cronbach's alpha was .95. Each of the three tools was also subjected to factor analytic procedures. Using criterion measures of adherence behavior, we demonstrated that the instruments have adequate construct validity. The results from these scales show that they are reliable and valid; the high Cronbach's alpha for each scale was as expected. There is a need to identify appropriate self-efficacy scales and to assess existing tools. Given the continuing interest in self-efficacy, the possibility exists that there may be an already developed and tested behavior-specific self-efficacy measure available for use.

See more of Measurement: Design and Development of Instrumentation to Assess Patient Attitudes and Beliefs and to Collect Background Information
See more of The 17th International Nursing Research Congress Focusing on Evidence-Based Practice (19-22 July 2006)