Integrating five linear trend techniques into performance‐criteria‐based effect size measurements: Impressions and recommendations
Journal of Applied Behavior Analysis · 2026-08 · 59 (3)
DOI: 10.1002/jaba.70070 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Applied Behavior Analysis, online erschienen am 24.05.2026. Verfasst von Orhan Aydin. Beteiligte Einrichtung: Division of Autism Spectrum Disorder, Department of Special Education, Faculty of Educational Sciences, Ankara University, Ankara, Türkiye. Hauptthemen laut MEDLINE: Statistics as Topic. Die Arbeit nennt 88 Literaturangaben.
Abstract (Originalsprache)
Effect size measurements in single‐case experimental designs are essential for identifying evidence‐based practices; however, each effect size index has notable limitations. To address constraints of nonoverlap‐based indices, Aydin and Tanious (2022) introduced the performance‐criteria‐based effect size (PCES), which quantifies socially meaningful behavioral change. However, the original PCES for trend correction (PCES trend ) uses a basic split‐middle method. This study sought to enhance PCES trend measurements by incorporating five advanced linear trend‐fitting techniques. A total of 87 published graphs with therapeutic baseline trends were analyzed, comparing techniques using performance indicators such as mean absolute scaled error, coefficient of variation, and R 2 , along with visual judgments and ancillary outcomes (e.g., out‐of‐bounds forecasts and missing data imputation). Results revealed that ordinary least squares (OLS) provided the most accurate and reliable adjustments. Based on these findings, PCES calculations with OLS were incorporated into a user‐friendly Google Sheet tool to increase accessibility and facilitate applied use.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Division of Autism Spectrum Disorder, Department of Special Education, Faculty of Educational Sciences, Ankara University, Ankara, Türkiye.
ORCID-Profile: Orhan Aydin
Zitieren
APA
Aydin, O. (2026). Integrating five linear trend techniques into performance‐criteria‐based effect size measurements: Impressions and recommendations. Journal of Applied Behavior Analysis, 59(3). https://doi.org/10.1002/jaba.70070
Vancouver
Aydin O. Integrating five linear trend techniques into performance‐criteria‐based effect size measurements: Impressions and recommendations. Journal of Applied Behavior Analysis. 2026;59(3). doi:10.1002/jaba.70070
BibTeX
@article{Aydin2026,
author = {Aydin, Orhan},
title = {Integrating five linear trend techniques into performance‐criteria‐based effect size measurements: Impressions and recommendations},
journal = {Journal of Applied Behavior Analysis},
year = {2026},
volume = {59},
number = {3},
doi = {10.1002/jaba.70070}
}
EUVIMED