Accomplishing micro-macro aggregations with richly parametrized linear models

While many sociologists focus on contextual drivers of individual behavior (e.g. neighborhoods on individuals), there are few studies on the ways individual behaviors influence social environment. One reason for the dearth of empirical work supporting micro-to-macro links is that there is no method available within regression analysis to properly study this. We developed a conceptually reversed hierarchical linear model (HLM) that utilizes the ‘multiple membership’ variant of the HLM to incorporate an aggregation function into regression modeling. Previous studies examining micro-to-macro links either aggregated or disaggregated the data. These approaches obstruct the inherent aggregation problem, cannot separate micro- and macro-level residual variance, ignore dependencies among observations, and induce Type-I error. The proposed model overcomes these problems by making the aggregation problem explicit in the model formulation and including both a micro- and macro-level residual term. It is a theoretically and statistically sound solution to the study of micro-to-macro links with regression analysis.

  • The project has been presented at PAA 2018 and the International Multilevel Conference 2017

Leave a Reply

Your email address will not be published. Required fields are marked *