Abstract
This work protocol form part of a three-phase publication (Marin-Garcia, 2019). Its objective is to establish a work procedure to answer these questions: 1) in which journals have articles about Operations Management with Partial Least Squares (PLS) been published?; 2) Do the results of previous reviews on this topic still prevail based on the very limited set of journals that it have been conducted (and before substantial modifications were made to report methods of PLS-based studies)?; 3) Do recent articles fulfil report recommendations; 4) What kind of measurement model has been considered for the constructs most frequently used in the selected articles?; 5) What are the usual R2 values in the cross-sectional studies represented in the selected articles?; 6) Within what statistical power range do the relations analysed with PLS fall?The article summarises current recommendations to select the analysis procedures and to report the research works that have used structural equations with PLS. We believe that this is an excellent contribution for researchers because it helps to improve the analyses and reports that derive from using PLS to, thus, increase the probabilities of them being accepted in relevant journals.Another contribution made by the present work, apart from establishing the aforementioned protocol, is to provide a list of the recent articles about operations management that have used PLS and the coding procedure to conduct our systematic review (to be subsequently published).
Highlights
The article summarises current recommendations to select the analysis procedures and to report the research works that have used structural equations with Partial Least Squares (PLS)
Another contribution made by the present work, apart from establishing the aforementioned protocol, is to provide a list of the recent articles about operations management that have used PLS and the coding procedure to conduct our systematic review
Listado de146 referencias que usan PLS en la investigación en Dirección de Operaciones en revistas clasificadas en WOS en las subáreas de “engineering” u “operations research”, procedentes de los resultados de WOS2b tras el filtro manual por parte del primer autor
Summary
El método PLS consiste en una secuencia de regresiones múltiples que permiten estimar los pesos de los componentes de los constructos (cuando se llega a un nivel de convergencia predefinido) y los paths entre constructos exógeneos y endógenos (Esposito Vinzi et al, 2010; Felipe et al, 2017; Henseler et al, 2009). De modo que podemos construir modelos más parsimoniosos (Ringle et al, 2018), agrupando las relaciones de conjuntos de variables que tienen cierto sentido teórico conjunto (Grace & Bollen, 2008), y se pueden interpretar como una unidad, pero sin perder el análisis del efecto de cada una de ellas por separado. Hasta la fecha se han identificado tres tipos de indicadores que son usados en diferentes modelos de medida en los constructos (nosotros usaremos constructos como una forma de referirnos indistintamente a variables latentes o variables emergentes) (Figura 6): effect (o reflective) indicator (que son usados en constructos common factor), causal indicator y composite indicator (que da origen a constructos composites) (Bollen & Bauldry, 2011; Grace & Bollen, 2008; Henseler et al, 2016b). Los diseños humanos, ya sean prácticas empresariales, modelos de gestión, índices de rendimiento o incluso competencias personales, al ser un objeto creado a partir de elementos, pueden tener una modelización adecuada con composites indicators, independientemente de que admita también otras modelizaciones, de modo que su ajuste a los datos debe ser confirmado por un análisis pertinente (Henseler, 2015; Henseler, 2017a)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
