Assessing Consistency in Single-Case Alternation Designs.
Use the free web app to quantify consistency of effect magnitude across blocks in alternation designs instead of relying on visual inspection alone.
01Research in Context
What this study did
Manolov et al. (2021) built a free web app. It scores how steady the effect size stays across blocks in alternation designs.
You paste in your data. The app gives you a consistency number and a color graph. No math by hand.
What they found
The paper shows the tool, not new client data. It explains how to read the output and what high or low consistency looks like.
How this fits with other research
Spiegel et al. (2023) took the idea further. They used the same logic on 460 ABAB graphs and set CONDAP cut-offs for "very high" to "very low" consistency.
Coon et al. (2018) looked at a different issue: only half of multiple-baseline graphs shaded concurrent phases. Both papers push for clearer single-case pictures.
Bachman et al. (1988) warned that leaving single-subject designs hurts progress. Rumen’s tool answers that call by making alternation data easier to trust.
Why it matters
Visual inspection alone can miss drift in effect size. The app gives you a number you can track and report. Try it on your next alternation design to spot unstable effects early and adjust the intervention before the graph is finished.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Upload last week’s alternation data to the app, note the consistency score, and decide if you need to tweak the contingency.
02At a glance
03Original abstract
Consistency is one of the crucial single-case data aspects that are expected to be assessed visually, when evaluating the presence of an intervention effect. Complementarily to visual inspection, there have been recent proposals for quantifying the consistency of data patterns in similar phases and the consistency of effects for reversal, multiple-baseline, and changing criterion designs. The current text continues this line of research by focusing on alternation designs using block randomization. Specifically, three types of consistency are discussed: consistency of superiority of one condition over another, consistency of the average level across blocks, and consistency in the magnitude of the effect across blocks. The focus is put especially on the latter type of consistency, which is quantified on the basis of partitioning the variance, as attributed to the intervention, to the blocking factor or remaining as residual (including the interaction between the intervention and the blocks). Several illustrations with real and fictitious data are provided in order to make clear the meaning of the quantification proposed. Moreover, specific graphical representations are recommend for complementing the numerical assessment of consistency. A freely available user-friendly webpage is developed for implementing the proposal.
Behavior modification, 2021 · doi:10.1177/0145445520923990