Data fro...

The purpose of this paper is to provide a brief review of multilevel modelling (MLM), also called hierarchical linear modelling (HLM), and to present a step-by-step tutorial on how to perform a crossed random effects model (CREM) analysis. The first part provides an overview of how hierarchical data have been analyzed in the past and how they are being analyzed presently. It then focuses on how these types of data have been dealt with in psycholinguistic research. It concludes with an overview of the steps involved in CREM, a form of MLM used for psycholinguistics data. The second part includes a tutorial demonstrating how to conduct a CREM analysis in SPSS, using the following steps: 1) clarify your research question, 2) determine if CREM is necessary, 3) choose an estimation method, 4) build your model, and 5) estimate the model’s effect size. A short example on how to report CREM results in a scholarly article is also included. Many statistical techniques, such as analysis of variance (ANOVA) and ordinary least-squares (OLS) multiple regression, assume that observations are not correlated with one another. However, this is not always the case. Within many areas of psychology, measurements are not full