Multivariate evaluation of health attitudes and behaviors: development and validation of a method for health promotion research

Prev Med. 1985 Jan;14(1):34-54. doi: 10.1016/0091-7435(85)90019-2.


Many self-reported measures of health status, attitudes, and knowledge used by clinicians and researchers suffer from a variety of shortcomings, including limited empirical justification, excessive complexity, assessments of nonmodifiable historic or hereditary factors, and limited utility for public health program planning. The Health Attitudes and Behavior Scale is an instrument designed to overcome many of these shortcomings and direct public health professionals toward more efficacious interventions. One hundred sixty-four subjects responded to true-false and Likert scale items related to health beliefs, attitudes, and behavioral practices. A principal component analysis of the Likert items yielded six components, tentatively labeled (a) Lack of Social Support, (b) Hurdles to Health, (c) Health Attitudes/Weight Concerns, (d) Positive Environment, (e) Disease Concerns, and (f) Time and Work Pressure. Analyses of these scales using coefficient alpha indicated adequate internal consistency for each of them. These scales were then related to demographic variables of age, educational level, sex, and marital status. Next, these scales were correlated with individual true-false items reflecting self-reported behavioral practices or health histories. A component analysis of the true-false items yielded four principal components labeled (a) Organizational Health Concerns, (b) Smoking and Exercise Concerns, (c) Coronary and Weight Concerns, and (d) Blood Pressure and Risk Factor Programs. Finally, a cluster analysis yielded six typical profiles reflecting different levels of the original six components. Of interest is the fact that the Lack of Social Support and Time and Work Pressure scales had a very limited integration into the overall patterns. The assumption that organizational and environmental factors can have an important impact on health was supported. The need for further research in this area is also discussed.

Publication types

  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Attitude to Health*
  • Behavior*
  • Blood Pressure
  • Body Weight
  • Coronary Disease / etiology
  • Coronary Disease / prevention & control*
  • Female
  • Health Promotion*
  • Health Status
  • Humans
  • Male
  • Middle Aged
  • Personality Assessment
  • Research
  • Rhode Island
  • Risk