Commentary on Slocum et al. (2022): Additional Considerations for Evaluating Experimental Control.

Abstract

In the target article, Slocum et al. (2022) suggested that nonconcurrent multiple baseline designs can provide internal validity comparable to concurrent multiple baseline designs. We provide further support for this assertion; however, we highlight additional considerations for determining the relative strength of each design. We advocate for a more nuanced approach to evaluating design strength and less reliance on strict adherence to a specific set of rules because the details of the design only matter insofar as they help researchers convince others that the results are valid and accurate. We provide further support for Slocum et al.'s argument by emphasizing the relatively low probability that within-tier comparisons would fail to identify confounds. We also extend this logic to suggest that staggering implementation of the independent variable across tiers may be an unnecessary design feature in certain cases. In addition, we provide an argument that nonconcurrent multiple baseline designs may provide verification within baseline logic contrary to arguments made by previous researchers. Despite our general support for Slocum et al.'s assertions and our advocacy for more nuanced approaches to determining the strength of experimental designs, we urge experimenters to consider the perspectives of researchers from other fields who may favor concurrent multiple-baseline designs and suggest that using concurrent multiple-baseline designs when feasible may foster dissemination of behavior analytic research.