The purpose of this research was to construct instruments to measure two personality variables that might explain individual differences in preventive behaviors for heart disease and to test the instruments for reliability and validity. The instruments were constructed to measure three concepts: health value orientations and perceived barriers and benefits to undertaking preventive behaviors believed to reduce one's risk for coronary artery disease. The concepts were components of the Preventive Behavior Model which was derived from social learning theory. Seventy-six subjects who had undergone a health screening program were tested. The testing session consisted of completing a Demographic Data Form which obtained information on smoking and exercise behaviors, the Marlowe-Crowne Social Desirability scale, the Likert Barrier/Benefit scale (BASBES), and Likert Value Orientation scale (VOS), and five magnitude estimation scales. Reliability testing included both stability and internal consistency. Test-retest coefficients for the Barrier/Benefit subscales were .52 and .71, and ranged from .32 to .68 for the Value Orientation subscale variations. Test-retest coefficients for the magnitude subscales ranged from .67 to .90, indicating the scaling technique obtained more stable results. Both alpha and theta coefficients were calculated to estimate internal consistency of the Likert scales. Alpha coefficients were .81 and .80 for the Barrier and Benefit subscales respectively and theta was .82 and .81 respectively. Theta coefficients ranged from .46 to .72 for the Value Orientation subscale variations while alphas were much lower, evidence that the items were not parallel. Construct validity was estimated by principal components factor analysis and predictive modeling. The orthogonal solution for the Barrier subscale revealed two components of the concept were being tapped. One factor resulted from rotation of the Benefit subscale. Factor analysis results suggested that many of the items on the VOS were not tapping the concepts as theoretically predicted. Although rotation of the factors resulted in one meaningful factor for each subscale variation, only 40 to 80 percent of the items were loading on the factors as hypothesized. Predictive modeling using stepwise regression analysis indicated that six of the 12 variables tested were impacting on one preventive behavior (exercise) as theorized.
| Identifer | oai:union.ndltd.org:arizona.edu/oai:arizona.openrepository.com:10150/187692 |
| Date | January 1982 |
| Creators | MURDAUGH, CAROLYN LUCILLE. |
| Publisher | The University of Arizona. |
| Source Sets | University of Arizona |
| Language | English |
| Detected Language | English |
| Type | text, Dissertation-Reproduction (electronic) |
| Rights | Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author. |
Page generated in 0.0024 seconds