Indexed in

License and use

Citations

Cited 285 times in

Cited 230 times in

Cited 71 times in

Altmetrics

Grant support

The codes used in the current study are available at an Open Science Framework repository, for reproducibility purposes: https://osf.io/e9f2c/?view_only=3732b311ef304b1793ee92613dcb0fe7.Research reported in this publication was supported by the National Institute on Aging of the National Institutes of Health under award R01AG024270. Luis Eduardo Garrido is supported by Grant 2018-2019-1D2-085 from the Fondo Nacional de Innovacion y Desarrollo Cientifico y Tecnologico (FONDOCYT) of the Dominican Republic. Jotheeswaran Amuthavalli Thiyagarajan and Ritu Sadana are staff members of the World Health Organization. All listed authors alone are responsible for the views expressed in this publication and they do not necessarily represent the decisions, policy, or views of the World Health Organization.

Analysis of institutional authors

Share

Investigating the Performance of Exploratory Graph Analysis and Traditional Techniques to Identify the Number of Latent Factors: A Simulation and Tutorial

Publicated to:PSYCHOLOGICAL METHODS. 25 (3): 292-320 - 2019-12-01 25(3), DOI: 10.1037/met0000255

Authors: Golino, Hudson; Shi, Dingjing; Christensen, Alexander P.; Eduardo Garrido, Luis; Dolores Nieto, Maria; Sadana, Ritu; Thiyagarajan, Jotheeswaran Amuthavalli; Martinez-Molina, Agustin;

Affiliations

Abstract

Exploratory graph analysis (EGA) is a new technique that was recently proposed within the framework of network psychometrics to estimate the number of factors underlying multivariate data. Unlike other methods, EGA produces a visual guide-network plot-that not only indicates the number of dimensions to retain, but also which items cluster together and their level of association. Although previous studies have found EGA to be superior to traditional methods, they are limited in the conditions considered. These issues are addressed through an extensive simulation study that incorporates a wide range of plausible structures that may be found in practice, including continuous and dichotomous data, and unidimensional and multidimensional structures. Additionally, two new EGA techniques are presented: one that extends EGA to also deal with unidimensional structures, and the other based on the triangulated maximally filtered graph approach (EGAtmfg). Both EGA techniques are compared with 5 widely used factor analytic techniques. Overall, EGA and EGAtmfg are found to perform as well as the most accurate traditional method, parallel analysis, and to produce the best large-sample properties of all the methods evaluated. To facilitate the use and application of EGA, we present a straightforward R tutorial on how to apply and interpret EGA, using scores from a well-known psychological instrument: the Marlowe-Crowne Social Desirability Scale. Translational Abstract Understanding the structure and composition of data is an important undertaking for a wide range of scientific domains. An initial step in this endeavor is to determine how the data can be summarized into a smaller set of meaningful variables (i.e., dimensions). In this article, we extend a state-of-the-art network science approach, called exploratory graph analysis (EGA), used to identify the dimensions that exist in multivariate data. Using Monte Carlo methods, we compared EGA with several traditional eigenvalue-based approaches that are commonly used in the psychological literature including parallel analysis. Additionally, the simulation study evaluated the performance of new variants of the EGA method and considered a wider set of realistic conditions, such as unidimensional structures and variables of continuous and categorical levels of measurement. We found that EGA performed as well as or better than the most accurate traditional method (i.e., parallel analysis). Importantly, EGA offers a few advantages over traditional methods: (a) it provides an intuitive visual representation of the results, (b) this representation offers a more complex understanding of the data's structure, and (c) the algorithm is deterministic meaning there are fewer researcher degrees of freedom. In sum, our study demonstrates that EGA can accurately identify the underlying structure of multivariate data, while retaining the complexity of the data's structure. This implies that researchers can meaningfully summarize their data without sacrificing the finer details.

Keywords