Publication:
Getting the best out of single case data

Thumbnail Image
Files
gre-sced.pdf(138 KB)
Date
2025
relationships.isAuthorOfPublication
relationships.isSecondaryAuthorOf
relationships.isDirectorOf
Authors
Leppink, Jimmie
item.page.secondaryauthor
item.page.director
Publisher
Universidad de Murcia: servicio de publicaciones
publication.page.editor
publication.page.department
Sin departamento asociado
DOI
https://doi.org/10.6018/edumed.664271
item.page.type
info:eu-repo/semantics/article
Description
Abstract
Given its potential for personalized healthcare and education, single case designs(SCDs) (1-4) – in which the same cases are measured repeatedly on the same outcome(s) ofinterest – has enjoyed increasing interest over time. In a publication in Volume 3 of thisjournal (5), three key statistical methods for SCDs are compared and contrasted:randomization tests (4), the Bayesian percentage of all non-overlapping data or PAND-B (5),and a time series regression (TSR) method (5). Although PAND-B is the only of thesemethods that allows for the incorporation of knowledge from theory or previous researchand does not require frequently violated assumptions that invalidate TSR (5), recent work(6) demonstrates that PAND-B does not fully account for order information in ordinal orquantitative outcomes (e.g., ‘3’ is not only different from ‘1’ or ‘2’; ‘3’ is also higher than ‘2’,and ‘2’ is higher than ‘1’) and returns invalid estimates when the two conditions comparedhave unequal sample sizes. To address these shortcomings while retaining the benefits ofbeing useful for all levels of measurement including nominal (e.g., different shoe colors orqualitatively different emotions that cannot undisputedly be categorized on a dimensionlike ‘better’ or ‘worse’) and incorporating knowledge from theory or previous research, ageneralized relative effect (GRE) statistic was developed (6), which is based on Brunner-Munzel’s relative effect (RE) statistic (7). Although the GRE introductory paper provides astep-by-step tutorial for how to obtain GRE point and interval estimates with basic R (8)code and interpret the outcomes, that paper focuses on two larger samples of participants.Therefore, the remainder of this paper uses the data from the aforementioned publicationin Volume 3 of this journal (5) to provide an example of GRE in a single context.
publication.page.subject
Citation
Leppink, J. (2025). Aprovechar al máximo los datos de casos únicos. Revista Española De Educación Médica, 6(3). https://doi.org/10.6018/edumed.664271
URI
http://hdl.handle.net/10201/199429
item.page.embargo
Collections
Revista Española de Educación Médica Vol. 6 Nº 3 (2025)
Ir a Estadísticas