Multilevel modeling and its application in counseling psychology research

Multilevel modeling (MLM) should be used when a researcher has collected hi erarchical data. For example, when a researcher investigates an outcome var iable (e.g., depression) with several clients drawn from different clinicia ns, the data set has a hierarchical structure. Herein, we describe the use of MLM in counseling research. The goals include the following: (a) to spec ify research contexts where MLM may be applied (b) to describe how to condu ct data analyses using MLM, and (c) to highlight key statistical and design issues encountered when analyzing hierarchical data. We also highlight how MLM can be used (a) to provide valid statistical inference in the presence of hierarchical data structure. (b) to separate the within-group effects f rom between-group, effects for predictor variables, and (c) to study the in teractions among predictor variables drawn from different levels (e.g., var iables drawn from both clients and their clinicians).