Regression discontinuity (RD) designs are increasingly used by researchers to obtain unbiased estimates of the effects of education-related interventions. These designs are applicable when a continuous “scoring” rule is used to assign the intervention to study units (for example, school districts, schools, or students). Units with scores below a pre-set cutoff value are assigned to the treatment group and units with scores above the cutoff value are assigned to the comparison group, or vice versa. For example, students may be assigned to a summer school program if they score below a preset point on a standardized test, or schools may be awarded a grant based on their score on an application.
Under an RD design, the effect on an intervention can be estimated as the difference in mean outcomes between treatment and comparison group units, adjusting statistically for the relationship between the outcomes and the variable used to assign units to the intervention, typically referred to as the “forcing” or “assignment” variable. A regression line (or curve) is estimated for the treatment group and similarly for the comparison group, and the difference in average outcomes between these regression lines at the cutoff value of the forcing variable is the estimate of the effect of the intervention. Stated differently, an effect occurs if there is a “discontinuity” in the two regression lines at the cutoff. This estimate pertains to average treatment effects for units right at the cutoff. RD designs generate unbiased estimates of the effect of an intervention if (1) the relationship between the outcome and forcing variable can be modeled correctly and (2) the forcing variable was not manipulated to influence treatment assignments.
This document presents criteria under which RD designs Meet WWC Evidence Standards and Meet WWC Evidence Standards with Reservations.