In this paper we propose a way of analyzing data from a plaid square design using multilevel mixed models. In the case of normal outcomes, this can be seen as a generalization to ANOVA analysis, where covariates can be included and extensions to unbalanced designs can be considered. Furthermore, based on the analysis on mixed models, the analysis of non-normal data can be considered, although fitting these models using existing software might prove a real challenge.
展开▼