This journal article format dissertation examined aspects of survey research methodology in health education. In the first study, the author examined articles published in Health Education and Behavior, Health Education Journal, Health Education Research, and International Electronic Journal of Health Education to assess if authors report survey instrument characteristics and results of psychometric property tests for data collected with survey instruments. In the second study, the author examined the validity and reliability of data collected from 1,992 Indiana middle and high school students with the Adolescent Health Risk Behavior Survey (AHRBS) instrument. The AHRBS instrument was created using the Biopsychosocial Model (BPSM) theoretical framework and investigates the relationships and influences of adolescents? intrapersonal and normative perceptions on alcohol, tobacco, and other drug (ATOD) use. In the third study, the author used reliable measures for the data to conduct mediation analyses to examine the effects of adolescent perceptions of their social environment, such as perceived peer disapproval, perceived parental disapproval, and perceived peer behavior, on adolescent lifetime inhalant use in the presence of the theorized mediator variable, characteristics of the peer group.
The author concluded the following: (1) published articles in the four health education journals inconsistently reported survey instrument characteristics or results of psychometric property testing for the data collected with survey instruments; (2) systematically eliminating items due to their limited contribution to scale reliability for these data using exploratory factor analyses, confirmatory factor analyses, and calculating internal consistency reliability shortened the AHRBS instrument by 41.18% and improved the reliability of measures for these data; and (3) the effects of perceived peer disapproval, perceived parental disapproval, and perceived peer behavior on adolescent lifetime inhalant use were significantly mediated by characteristics of the peer group as theorized by the BPSM.
Findings of this dissertation have implications for the field of health education. First, survey instrument characteristics and internal reliability coefficients for data collected with instrument scales should be reported in published manuscripts. Second, researchers should examine the validity and reliability of data collected with survey instruments. Third, only reliable measures for the data should be used in statistical analyses.
| Identifer | oai:union.ndltd.org:tamu.edu/oai:repository.tamu.edu:1969.1/ETD-TAMU-2008-12-174 |
| Date | 14 January 2010 |
| Creators | Smith, Matthew L. |
| Contributors | McKyer, E. Lisako J. |
| Source Sets | Texas A and M University |
| Language | en_US |
| Detected Language | English |
| Type | Book, Thesis, Electronic Dissertation |
| Format | application/pdf |
Page generated in 0.0094 seconds