Although multilevel modeling (MLM) has been suggested as one of the promising approaches to synthesizing SCED data, applications of MLMs in the SCED context are rare due to some challenges faced by applied researchers. One of the potential challenges of using this approach is the lack of user-friendly software or interface to conduct this analysis. Although there are several R packages available for general MLM analysis such as nlme or lme4, all of the current packages are developed for large sample-based data, as a result, essential functions to handle complex SCED data (e.g., small sample adjusted degrees of freedom methods, autocorrelated error) are often missing. This project aimed to develop a survey to identify the challenges of using quantified effect size measures and specific needs in the development of R package specifically tailored for SCED and applied to a web-based effect size calculator for MLM effect sizes of SCED.

Survey results showed that 56% of researchers use either R packages or online calculators to estimate effect sizes in SCED studies. However, 67% reported dissatisfaction with existing software and effect size measurement tools. Researchers also identified a lack of user-friendly calculators and software tutorials (30%) as a major obstacle to obtaining effect sizes. To address these challenges, we developed the lmeSCED R package and a companion Shiny application. The Shiny application provides a more accessible interface for researchers. In addition, the lmeSCED package addresses autocorrelation through GSL transformation and incorporates a small-sample adjustment method. It also implements a boundary-corrected restricted likelihood-ratio test and parametric bootstrapping procedures for variance component inference.

Li, C., Baek, E., & Luo, W. (2026). Generalized least squares transformation for single-case experimental design: Introducing the R package lmeSCED. Behavior Research Methods, 58, Article 72. https://doi.org/10.3758/s13428-025-02936-4

Project Period: September 2023 – August 2024

PI: Eunkyeng Baek, Ph.D.

Funding Source: College of Education and Human Development, Texas A&M University