Increasing Equivalence in Small-Sample Place-Based Experiments

In recent years, places (such as small geographic locations, schools, classrooms, and clinics) rather than people have come into focus as the unit of analysis in randomized controlled trials (RCTs).  Places can be both the physical location and the organizational unit for human behavior (Boruch, Weisburd, and Berk 2010).  As such, many social reform interventions involve larger units of organization, and often policies developed by or for agencies such as police departments and schools are not always hypothesized to work directly at the level of the individual person (Boruch, et al. 2004; Campbell, 1969).  Influential place-based experiments have been conducted in the fields of health care  (e.g., Donner and Klar 2000; Grimshaw et al. 2005; Sikkema 2005; Imai, King, and Nall 2009) , social welfare (Bloom and Riccio 2005), education (Flay and Collins 2005; Parker and Teruel 2005), and criminal justice (e.g., Braga and Bond 2008; Braga et al, 1999; Sherman and Weisburd 1995; Taylor, Koper, and Woods 2011; Weisburd and Green 1995; Weisburd, Morris, and Ready 2008).