Public program analysis

Description:

Abstract: A multivariate categorical data analysis technique, termed weighted-least-squares (WLS), is applied to program analysis in health and public affairs for public health program (an other public affairs programs) planners. Program analysis encompasses the planning and prediction, assessment, and evaluation of programs. The WLS approach allows practitioners to analyze a wide range of problems, in contrast to conventional program analysis approaches in which different techniques are needed for each new problem type encountered. The WLS approach also produces estimates that are "best asymptotic normal", yielding an optimal approach to obtaining statistical estimates for large samples. Use of the WLS approach requires, at minimum, a knowledge of the analysis of variance (AOV). An introduction to matrices, and the use of matrices and the linear model in AOV situations is included for the statistically-inexperienced reader. (wz).

subjects: Statistical methods,  Least squares,  Multivariate analysis,  Evaluation research (Social action programs),  Public health,  Evaluation,  Dados categoricos,  Gesundheitswesen,  Statistik,  Statistische Analyse,  Evaluation Studies,  Statistics,  Evaluation Studies as Topic,  Least-Squares Analysis,  Statistics as Topic