Single-subject research encompasses a collection of research methods primarily utilized in the experimental analysis of behavior and applied behavior analysis. These methods are employed with both human and non-human participants, focusing intensely on one individual. The core principle involves tracking a participant's progress in a specific research topic over an extended period. This approach offers a unique advantage by allowing researchers to observe
changes within an individual across a significant timeframe, rather than comparing different people at various stages. While this research strategy can yield crucial data, particularly in fields like psychology, it has also been a subject of considerable debate regarding its overall effectiveness.
Core Methods and Their Applications
Several principal methods define single-subject research, each designed to address specific experimental questions. These include A-B-A-B designs, Multi-element designs, Multiple Baseline designs, Repeated acquisition designs, Brief experimental designs, and Combined designs. These methodologies are fundamental to the data collection and analytical framework of behavior analysis, a field characterized by its data-driven, inductive approach, and its disinclination towards hypothetico-deductive methods. The choice of experimental design is heavily influenced by the nature of the experimental question being asked.
Experimental questions typically fall into four basic categories: demonstration, comparison, parametric, and component. A demonstration question asks, "Does A cause or influence B?" A comparison question seeks to determine, "Does A1 or A2 cause or influence B more?" A parametric question investigates, "How much of A will cause how much change or influence on B?" Lastly, a component question explores, "Which part of A{1,2,3} - A1 or A2 or A3... - causes or influences B?" where A is composed of separable and testable parts. The A-B-A-B design, for instance, is particularly useful for demonstration questions.
The Power and Limitations of Reversal Designs
The reversal, or A-B-A, design stands out as one of the most powerful single-subject research designs. It demonstrates a strong reversal from a baseline phase ("A") to a treatment phase ("B") and then back to baseline. If the variable returns to its baseline measure when the treatment is removed, and then resumes its effects when the treatment is reapplied, researchers can have greater confidence in the treatment's efficacy. This cyclical pattern provides compelling evidence for a causal link between the treatment and the observed changes.
However, the application of reversal designs is not without its challenges and ethical considerations. Many interventions cannot be reversed, either for ethical reasons or practical ones. For example, it would be unethical to reverse treatments for self-injurious behavior or smoking, as doing so could cause harm or undo beneficial progress. Similarly, skills that have been learned cannot simply be "unlearned," making reversal impractical. Furthermore, ethical guidelines suggest that it may be inappropriate to end an experiment on a baseline measure if the treatment is highly beneficial and self-sustaining, especially when related to health. Researchers also have an ethical duty to maximize benefits and ensure that all participants, including those in control conditions, have access to the benefits of the research once data collection is complete.
Exploring Design Variants and Their Utility
Beyond the foundational A-B-A-B and A-B designs, several variants offer specialized utility. The A-B design, a two-part phase design, consists of a baseline ("A" phase) with no changes, followed by a treatment or intervention ("B") phase. While a change in this design might suggest an effect, it is susceptible to numerous competing hypotheses, making strong conclusions difficult. Variants of the A-B design aim to control for these competing hypotheses to allow for more robust conclusions. The A-B-C design, for example, is a variant that extends research questions to component, parametric, and comparative analyses.
Multi-element designs, also known as alternating-treatment designs, are employed to compare the effects of two treatments. These designs involve rapidly alternating two treatments and plotting correlated changes on a graph to facilitate comparison. They are typically used in single-subject research to accurately test multiple independent variables simultaneously. The multiple baseline design, first reported in 1960 for basic operant research and later applied to human experiments in the late 1960s, addresses practical and ethical issues associated with withdrawing successful treatments. In this design, two or more behaviors, people, or settings are plotted in a staggered graph. A change is introduced to one, but not the others, then to the second, and so on. Differential changes across these elements help strengthen what is essentially an A-B design by mitigating problematic competing hypotheses. This method yields a high amount of data, which can support hypotheses or provide insights before moving to group research projects. Other designs like repeated acquisition, brief experimental designs favored in applied settings, and combined designs, which merge multiple single-case designs, further expand the toolkit for single-subject researchers, allowing for more complex research questions to be addressed.
