Single-Subject Design

A single-subject design, which is also known as a single-case research design, provides a rigorous approach for documenting experimental effects. Single-case research has been used to (a) establish basic principles of behavior, (b) document the impact of specific interventions, and more recently (c) establish evidence-based practices. The defining feature of single-case research is the use of each participant (subject) as his or her own experimental control. This approach to research design arose from early work of researchers such as Burrhus F. Skinner and Werner Honig in the 1940s and 1950s focusing on behavior analysis, and it was codified in the seminal book Tactics of Scientific Research by Murray Sidman in 1960. Sidman defined in detail how the systematic study of individual participants over time could be used to test important experimental concepts and expand our fundamental understanding of human behavior.

Single-case methods have become a standard approach for conducting educational, behavioral, and psychological scholarship. Although single-case research is associated with behavioral psychology (e.g., Journal of Applied Behavior Analysis), single-case methods are now used to document advances across an array of fields including social-learning theory, medicine, social psychology, education, social work, pediatric psychology, and communication disorders.

The basic goal of single-case research designs is to evaluate the extent to which a causal (or functional) relation exists between introduction of an “intervention” (e.g., pace of instruction) and change in a specific dependent variable (e.g., reading performance). In most cases, a researcher conducting a single-case study wishes to establish that (a) prior to intervention a subject behaved in a consistent, clear manner that does not meet social expectations (e.g., the student read slowly or inaccurately, or the student engaged in high rates of problem behavior); (b) after the intervention the participant behaved in a manner that is both different and better than preintervention (e.g., the student read faster and/or more accurately, or the student behaved with less problem behavior and more positive behavior); and (c) this clearly measured change is unlikely to be the result of [Page 1387]anything other than the intervention (e.g., not normal development, not some other unmeasured intervention, and not illness or change in the personal life of the participant). It is this final assertion that the change in behavior was causally associated with introduction of the intervention that establishes the scientific credibility of single-case designs. Here, the defining features of single-case experimental research designs are reviewed, noting their contribution to a science of human behavior.