ISSN 1842-4562
Member of DOAJ

An Overview of Causal Directed Acyclic Graphs

Michael LEWIS


directed acyclic graphs, DAG, social work researchers, randomized controlled trials, RCT


Given the mission of social work to improve people’s lives by intervening in ways to enhance their well-being, social work researchers are interested in the causal effects of various types of social interventions (de Anda, 2007; Ohmer, and Jorr, 2006; Hawkins, 2006). Statisticians, econometricians, and other experts in quantitative methods tend to view randomized controlled trials (RCTs) as the “gold standard” when it comes to estimating causal effects, and there’s a key reason for this--RCTs are based on randomly assigning units to at least two different intervention groups. Random assignment tends to result in the groups being balanced on variables that may have a causal relationship with the outcome of interest other than the intervention which is the focus of a given study. At least one of these “balanced variables” may also be causally related to the intervention itself. “Balanced” means that the average values of these other variables are equal across the different groups.