ABSTRACT
This chapter provides readers with a template for conducting small agency-based studies, including writing up the findings to inform stakeholders. It discusses the components of a data analysis plan, a strategy for statistical analysis, and the central elements of evaluation reports. The key to both analysis and report writing is “making sense” of the data and this chapter will describe this process. The data analysis plan has five steps: specify the measurements, decide on the level of statistical significance and the methods for estimating clinical significance, make a plan for statistical analysis, devise a raw data examination and correction protocol, and develop a data-coding and entry protocol. The chapter describes assumes that evaluators will code and enter data themselves. It shows that the best way to analyze data is from the univariate to the bivariate to the multivariate. By examining the univariate characteristics of variables, an evaluator determines whether the statistics selected in the data analysis plan are appropriate.
