Mastering the Use of Control Variables: the Hierarchical Iterative Control (HIC) Approach

Abstract

There has been growing criticism of the established practice of automatically including control variables into analyses, especially with survey studies. Several authors have explained the pitfalls of improper use and have provided some best practice advice. I build upon this foundation in suggesting a programmatic approach to the use of control variables that can provide evidence to support or refute feasible explanations for why two or more variables are related. The hierarchical iterative control (HIC) approach begins by establishing a connection between two or more variables and then hierarchically adds control variables to rule in or out their possible influence. The HIC approach involves conducting a series of studies to iteratively test relationships among target variables, utilizing a variety of control variable strategies involving multiple methods. A 7-step programmatic approach is described beginning with development of the research question and background literature review and then conducting empirical tests in a hierarchical (within a study) and iterative (across studies) manner.