(M)ARI – A comparison of quantitative verification of formative items for index formation using R and SmartPLS
Empirische Pädagogik · 15.04.2026 · 40 (1) · 96-125
DOI: 10.62350/eoto4074 ↗Kurz gefasst
Fachartikel in der Zeitschrift Empirische Pädagogik, online erschienen am 15.04.2026. Verfasst von Hannah Lichtenberg.
Abstract (Originalsprache)
The MARI method is an instrument for data-driven index formation for the verification and creation of formative measurement models. This paper is dedicated to the quantitative procedures of the method. The model is initially verified based on quantitative analyses of the items using descriptive observations, a review of correlations using CTA-PLS or CFA, a review of multicollinearity, and a quantitative auxiliary analysis for content validity. The model is validated using linear regressions or PLS-SEM. Finally, there is a possibility for index formation. The individual quantitative steps of the MARI method are explained in terms of content and methodology, and their practical implementation is outlined. The method is applied by comparing the statistics programme SmartPLS and the programming language R. To illustrate the MARI method, a data set on job satisfaction among daycare centre employees is analysed. The two programmes are characterised by different approaches, but the method can be implemented without any problems using either programme.
Zitieren
APA
Lichtenberg, H. (2026). (M)ARI – A comparison of quantitative verification of formative items for index formation using R and SmartPLS. Empirische Pädagogik, 40(1), 96-125. https://doi.org/10.62350/eoto4074
Vancouver
Lichtenberg H. (M)ARI – A comparison of quantitative verification of formative items for index formation using R and SmartPLS. Empirische Pädagogik. 2026;40(1):96-125. doi:10.62350/eoto4074
BibTeX
@article{Lichtenberg2026,
author = {Lichtenberg, Hannah},
title = {(M)ARI – A comparison of quantitative verification of formative items for index formation using R and SmartPLS},
journal = {Empirische Pädagogik},
year = {2026},
volume = {40},
number = {1},
pages = {96--125},
doi = {10.62350/eoto4074}
}
EUVIMED