Examining and Enhancing the Methodological Quality of Nonconcurrent Multiple-Baseline Designs

Kratochwill et al. (2022) looked at nonconcurrent multiple-baseline designs. These are studies where each participant starts at a different time.

The authors wrote a how-to paper. They explained ways to make these designs stronger without running them all at once.

The paper says you can still guard against history effects. Just add random start points and repeat tiers.

They admit concurrent designs are safer. Yet with their tweaks, nonconcurrent ones can pass peer review.

Slocum et al. (2022) almost disagrees. They say nonconcurrent designs can match concurrent ones if you pile enough within-tier replications. Kratochwill adds random timing on top.

Petursdottir et al. (2018) came first. They mapped all validity threats for single-case work. Kratochwill zooms in on one threat and offers a fix.

Barnard‐Brak et al. (2021) tackles a different problem: inflated effect sizes. Both papers push the same message—tighten your method or your numbers will fool you.

School teams often can’t start everyone on the same day. If you must stagger, now you have a checklist: randomize start points, replicate tiers, and document each step. Reviewers will trust your data, and you can keep using practical start dates.