ABSTRACT
This chapter addresses whether groups are significantly different on some dependent variable. The variation on this theme is how we can achieve greater power and precision in assessing differences across groups by incorporating information about individual differences in our participants into our models. The chapter considers how best to make use of information we might obtain about the individual differences among participants. It considers the approaches to handling concomitant variables that are represented by both analysis of covariance (ANCOVA) and blocking. The chapter begins with a consideration of analysis methods that treat the concomitant variable as a continuous variable. The primary data-analysis method of ANCOVA is compared with other related approaches, namely the analysis of change scores and the analysis of residuals. The chapter considers the methods of analyzing blocked designs and includes a discussion of issues that arise when the blocking is carried out after, rather than before, the study is run.
