Multiple baseline designs: the use of a single-case experimental design in literacy research

Reading Improvement, Winter, 2004 by Richael Barger-Anderson, Joseph W. Domaracki, Nedra Kearney-Vakulick, Richard M. Kubina, Jr.

Experimental Standards

Internal and External Validity. As with group designs, internal and external validity are important issues to be considered when using single-case designs. When changes occur that are attributable to the effects of the independent variable, the study has internal validity (Poling & Grossett, 1986). Internal validity of a single-case design is considered acceptable if an intervention is reliably associated with higher response levels while also revealing sound experimental control (Neuman & McCormick, 1995). In the case of a multiple baseline design, the changes occuring in each phase arise from the systematic application of the independent variable, not some extraneous variable.

External validity refers to the degree to which the results from a study can be generalized to other groups or settings (Gay & Airasian, 2000). Regardless of the type of experimental design used, single-case or group design, a number of factors encompass external validity (Poling & Grossett, 1986). Neuman and McCormick (1995) suggest that the best ways to attend to the issues of external validity include: 1) providing a rich and detailed description of the setting and the intervention, 2) detailing the measures, and 3) generalizing the results to a particular theory.

Control. Gay and Airasian (2000) explain how control is a primary characteristic of experimental studies: "Direct manipulation by the researcher of at least one independent variable is the one single characteristic that differentiates experimental research from other types of research. Control refers to the researcher's efforts to remove the influence of any extraneous variable (other than the independent variable itself) that might affect scores on the dependent variable" (p. 370). Dermer and Hoch (1999) propose that the single-case researcher views control as a way to discard variability caused by anything other than the treatment or independent variable. Control should continue until the researcher can visually determine an effect of the treatment. Single-case research uses control procedures rather than control groups (Good, 2000). Each student, then, serves as her or his own control (Gay & Airasian, 2000; Wolery & Gast, 2000).

Control can be difficult to achieve in a research design, particularly when dealing with people (Gay & Airasian, 2000). Control is even more challenging to achieve in the multiple baseline design when different interventions are offered sequentially. Counterbalancing is a possible way to control for order effect when using two or more interventions in multiple baseline design research. "The counterbalancing of treatments permits a comparison of the effectiveness of the two instructional methods" (Domaracki, 1987, p. 57).

Counterbalancing is completed in a study to address the concern of order effects. The first subject receives a treatment, a second subject receives a different treatment, then once the subjects have shown improvements, the treatments are reversed. Counterbalancing is an attempt to control for order effects (McReynolds & Kearns, 1983).