| The purpose of this dissertation is to model biosocial determinants of group-directed altruistic behavior -- exploring the nomological net around it. To do this a study will be presented to determine existing associations among various biological and social predictors and test a life-history-derived causal cascade using a partially exploratory and partially confirmatory statistical technique called Sequential Canonical Analysis to ultimately predict living-donor, non-directed kidney donations (NDKD). Toward that end, some important methodological considerations first need to be discussed. The first consideration revolves around the level of analysis and how this frames the cascade model and its interpretation. Following a general discussion, an exercise in some of the general principles is provided -- investigating the higher-order factor structure of the Big-5 personality constructs across two levels of analysis. The second consideration is the use of unit-weighted factor scores and their appropriateness. Following the theoretical discussion, a demonstration is provided -- deriving an estimate of genetic relatedness from a set of heterogeneous data sets. Once the methodological considerations have been discussed, the primary cascade model is presented in two parts: 1) the measurement model -- operationalizing the measures incorporated into 2) the structural model -- testing the proposed causal cascade using Sequential Canonical Analysis. A discussion follows in which the results are summarized, limitations are articulated, and further research directions are explored. |
