Project data analysis
Graham Currell · Oxford University Press eBooks · 2015
This chapter focuses on project data analysis. It begins by detailing the initial steps in reviewing data and objectives, understanding the source of data uncertainties, and transforming the data into a format suitable for software analysis. For electronic records, the original data should be saved in a secure memory store before copying the data into a new file, and all subsequent analysis should only be carried out on the duplicate data set. The chapter then presents examples of methods to identify experimental characteristics that are suitable for analysis. It also reviews the techniques and issues involved in transforming the data into a format required to match the analytical technique. Finally, the chapter discusses the relevance of normality and equality of variance for many analytical techniques and develops the approach, tests, and transformations that may be involved.