Document Type



Two- and three-level hierarchical linear modeling is introduced and demonstrated as a meta-analysis technique to estimate treatment effects across cases and studies. Multilevel modeling techniques have been empirically validated and are promising (Moeyaert, Ugille, Ferron, Beretvas, & Van den Noortgate, 2013a, 2013b). However, detailed instruction is needed to bridge the gap between methodologists and applied researchers. An empirical illustration estimating the effectiveness of choice-making interventions on decreasing students with special needs’ problem behaviors is discussed. Step-by-step formulations of equations and R code for applied researchers are provided in order for researchers to conduct multilevel analyses using their own SCED data.