Dyadic Data Analysis Using Multilevel Modeling
Recently Kenny, Kashy, and Cook (2006) have extensively discussed the analysis of dyadic data, and MLM is an important tool in many of the analyses they describe. In this chapter, we review and extend their discussion. We begin with key definitions in dyadic analysis concerning types of dyads, types of dyadic designs, and types of variables. We then consider the use of MLM for the three major types of dyadic designs. Finally, we consider the analysis of over-time data for one of those designs. Most dyadic researchers would benefit from reading the entire chapter even if they have an interest in only one of these designs. Topics that are discussed in one section (e.g., coding of variables, computer syntax, and interpretation of results) are relevant for the other sections of the chapter.