Single-subject research, a cornerstone of experimental analysis of behavior and applied behavior analysis, is fundamentally guided by the experimental questions it seeks to answer. This research strategy, which meticulously tracks the progress of a single participant over time, relies on a diverse set of methodologies. The selection of the appropriate research design is not arbitrary; rather, it is decisively determined by the nature of the experimental question at hand.
Understanding the different types of questions and their corresponding designs is crucial for effectively utilizing this powerful research approach.
Four Basic Types of Experimental Questions
There are four fundamental types of experimental questions that drive single-subject research, each probing a different aspect of a phenomenon. The first is a **demonstration** question, which asks, "Does A cause or influence B?" This type of question aims to establish a basic causal or influential relationship between an intervention (A) and an outcome (B). For instance, a researcher might want to demonstrate if a new teaching method influences a student's reading fluency.
The second type is a **comparison** question: "Does A1 or A2 cause or influence B more?" This question moves beyond simple demonstration to evaluate the relative effectiveness of two or more different interventions. An example would be comparing whether a visual aid or an auditory cue is more effective in influencing a specific behavior. The third category is a **parametric** question, which delves into the intensity or quantity: "How much of A will cause how much change or influence on B?" This seeks to understand the dose-response relationship, such as determining the optimal duration of an intervention to achieve a desired outcome. Finally, a **component** question asks, "Which part of A{1,2,3} - A1 or A2 or A3... - causes or influences B?" This is relevant when an intervention (A) is composed of multiple parts that can be separated and tested individually to identify the most effective elements.
Matching Questions with Research Designs
The specific experimental question dictates the most suitable single-subject research design. For instance, the A-B-A-B design is particularly well-suited for demonstration questions. This design establishes a baseline (A #1), introduces a treatment (B #1), returns to baseline (A #2) by removing the treatment, and then reintroduces the treatment (B #2). This cyclical pattern provides strong evidence for whether the treatment causes or influences the behavior.
For comparison questions, multi-element designs, also known as alternating-treatment designs, are often employed. These designs rapidly alternate two treatments and plot the correlated changes to facilitate a direct comparison of their effects. When dealing with component questions, the A-B-C design, a variant of the basic A-B design, allows for the extension of research questions to explore individual components. Similarly, the multiple baseline design, which staggers the introduction of an intervention across different behaviors, people, or settings, can help strengthen conclusions for what is essentially an A-B design by addressing problematic competing hypotheses. The careful alignment of experimental questions with appropriate research designs is fundamental to generating robust and meaningful data in single-subject research.
