Attitude surveys can serve as a basis of public interest and opinion inputs to National Park Service planning and management decision making. The initial objective of this study was to develop a method for scaling attitudes which alleviated problems of interpreting effects of two related components of attitude--beliefs and corresponding evaluations of those beliefs. In conjunction with this objective, visitor knowledge about and attitudes toward resource use and management in Shenandoah National Park were assessed. Finally, guidelines were formulated suggesting how to design an effective interpretive message taking into account relevant user characteristics, level of knowledge, and attitudes toward resource use and management in the national park. Considering summer months and weekend/weekday visitation patterns, a stratified random sample of days was selected to contact park visitors. On designated sampling days visitors were contacted at lodges, backcountry stations or campgrounds. Names and addresses were secured via an information release form. In September, 1978, individuals were mailed an attitude survey. A follow-up post card and two selective follow-up questionnaire mailings ensued. Overall response rate was 81 percent. The survey instrument obtained contributions of belief and affective components of attitude. In order to measure these components in a manner reflecting resource management objectives and allowing easy interpretation of attitude scores, a new modified matrix scaling technique was developed. A matrix of attitude scores was computed using an unipolar belief scale and bipolar affect scale. The computed using an unipolar belief scale and bipolar affect scale. The belief scale was also employed to determine visitor level of knowledge concerning resource use and management in Shenandoah National Park. The over-all matrix system was developed within the framework of Fishbein's (1963) expectancy value attitude theory. The scaling method lessened some of the problems created by using standard bipolar st.nnmative scaling. In addition, final scores derived from this matrix procedure were logically consistent, theoretically appropriate, and may be managerially relevant. Results of analyses of variance in overall belief and attitude scores suggested the utility of developing different interpretive messages about resource use and management for the three camping styleslodge, campground, and backcountry users. Attitude changes predicted via the expectancy value formulation indicated the importance of considering the distribution of visitor attitude scores before attempting development of an interpretive message designed to influence attitudes. / Ph. D.