Abstract
The purpose of this research is to contrast a traditional audience segmentation model that uses demographics and health evaluations against a model that uses these same variables plus media usage variables. The goal was to determine whether media usage variables—typically not used in health segmentation studies—add predictive power in determining health behaviors and attitudes. The results of the analysis showed an increase in the ability to predict health behaviors such as aspirin use, vitamin use, diet, and exercise, and suggest that there is predictive value for including media variables as part of the segmentation process. Implications for public health education and campaign planning are discussed.
The authors thank Dr. Robert Logan for his assistance with the survey. The authors are deeply indebted to the reviewers for their invaluable suggestions and insights.
This research was funded with a grant from the National Institute on Disability and Rehabilitation Research in conjunction with the Missouri Arthritis Rehabilitation and Research Training Center.
Notes
Education and income are used to show the central tendency of each cluster.
1. Education: 0 = never attended school/or only kindergarten; 1–11 = grades 1 through 11; 12 = grade 12 or GED (high school graduate); 13 = college 1 year to 3 years or some college or technical school; 14 = 4-year college degree; 15 = some graduate school but no degree; 16 = graduate degree.
2. Income: 0 = less than $10,000; 1 = at least $10,000 but under $20,000; 2 = at least $20,000 but under $30,000; 3 = at least $30,000 but under $50,000; 4 = at least $50,000 but under $75,000; 5 = at least $75,000 but under $100,000; 6 = at least $100,000 but under $150,000; 7 = $150,000 or more.
1. Aspirin use, vitamin use, special diet, exercise, and exercise class: 1 = yes, 2 = no.
2. Information search and self-reported changes to improve health: 1 = strongly disagree, 5 = strongly agree.
a p < .001;
b p < .01;
c p < .05;
d p < .10.
Note: Standard errors in parentheses.