Experimental Design and Validity Capella University

Experimental Design and Validity Overview This document is an academic paper that analyzes four different types of single-subject experimental designs used in Applied Behavior Analysis (ABA). For each design, the paper provides a relevant APA citation, details its strengths and limitations, and discusses its potential for Internal Validity, External Validity, and Social Validity. 1. Reversal Design (ABAB) Strength: Allows for rapid comparison and provides strong evidence of the correlation between the intervention and the dependent variables. It has high potential for internal validity due to the immediacy of effect and data consistency across similar phases. Limitation: Removing a treatment forces a return to the baseline. It may harm the subject where the intervention involves attaining skills or learning, as undoing learning can be difficult or impossible. Validity: Internal: Enhanced by using an ABAB design, procedural fidelity, and interobserver agreement. External: Limited due to a small study sample and the inclusion of participants with partial prior training. Social: The study goal addresses a significant problem, and use of the MotivAider® device enhanced the feasibility and acceptability of the process. 2. Multiple Baseline Design Strength: High validity and accuracy, minimizing the potential for internal and external influences. It's ideal for cases where the researcher does not expect to return to the baseline. Limitation: Requires intense data collection, especially when many participants or settings are involved. It's limited to research cases where a single type of treatment is required and cannot be applied with successive treatment interventions. Validity: Internal: Demonstrated by using a multiple-baseline design and a random number generator to reinforce internal control. External: The study setting simulated a real-world web-based learning platform, but the sample size might still limit generalizability. Social: The interventions (clinical decision model and outcomes) were based on the training needs of the setting, aligning with goals and objectives.