Effect size in single-case research: a review of nine nonoverlap techniques.
Pick one nonoverlap index early and use it for every graph so your effect sizes can play nicely in later meta-analyses.
01Research in Context
What this study did
The authors read every paper they could find on nine nonoverlap formulas. These formulas turn single-case graphs into one clean number. They lined up the math side-by-side to see which ones give similar answers.
No new clients were tested. The team simply unpacked the recipes for indices like PND, PEM, and Tau-U so meta-analysts can pick one with open eyes.
What they found
All nine indices claim to measure the same thing—how far data jump after treatment—but they rarely spit out the same value. Some ignore trend, others punish overlap, and a few reward steady gains.
The paper shows that switching indices mid-project can flip the size label from "small" to "large," so consistency matters more than the exact formula you choose.
How this fits with other research
Ferguson et al. (2022) and Ribeiro et al. (2024) both used alternating-treatments designs that could be scored with any of these nine indices. Their graphs would get different effect labels depending on the index picked, even though the visual story looks the same.
In-Lee et al. (2012) ran a meta-analysis on exercise for people with intellectual disability. They had to stick with one index to keep the average meaningful; this review explains why that discipline is required.
Kim et al. (2025) reviewed choice procedures but never pooled effect sizes. If they do a follow-up meta, this paper warns them to choose one nonoverlap method and stay with it.
Why it matters
When you write a single-case study, name the one nonoverlap index you will use before you see the data. Add it to your methods section just like you list your IOA formula. This small line of text keeps your effect size comparable to future studies and saves headaches if anyone runs a meta-analysis that includes your data.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Open your last single-case report, add a sentence that says which nonoverlap index you will use from baseline to follow-up, and stick with it.
02At a glance
03Original abstract
With rapid advances in the analysis of data from single-case research designs, the various behavior-change indices, that is, effect sizes, can be confusing. To reduce this confusion, nine effect-size indices are described and compared. Each of these indices examines data nonoverlap between phases. Similarities and differences, both conceptual and computational, are highlighted. Seven of the nine indices are applied to a sample of 200 published time series data sets, to examine their distributions. A generic meta-analytic method is presented for combining nonoverlap indices across multiple data series within complex designs.
Behavior modification, 2011 · doi:10.1177/0145445511399147