Exploratory Factor Analysis (EFA) in Quantitative Researches and Practical Considerations

Abstract

Explanatory factor analysis (EFA) is a multivariate statistical method frequently used in quantitative research and has begun to be used in many fields such as social sciences, health sciences and economics. With EFA, researchers focus on fewer items that explain the structure, instead of considering too many items that may be unimportant and carry out their studies by placing these items into meaningful categories (factors). However, for over sixty years, many researchers have made different recommendations about when and how to use EFA. Differences in these recommendations confuse the use of EFA. The main topics of discussion are sample size, number of items, item extraction methods, factor retention criteria, rotation methods and general applicability of the applied procedures. The abundance of these discussions and opinions in the literature makes it difficult for researchers to decide which procedures to follow in EFA. For this reason, it would be beneficial for researchers to gather different information about the general procedures (sample number, rotation methods, etc.) in the use of EFA. This paper aims to provide readers with an overview of what procedures to follow when implementing EFA and share practical information about the latest developments in methodological decisions in the EFA process. It is considered that the study will be an important guide for the researchers in the development of clear decision paths in the use of EFA, with the aspect of presenting the most up-to-date information collectively.

Exploring the psychometric properties of the digital citizenship scale among Indian students

“…MLE is considered appropriate for this study as it provides test statistics for evaluating the tenability of a particular model while maximizing the probability of obtaining the observed data. Moreover, CFA includes the maximum likelihood method; therefore, MLE allows cross-validation of EFA and CFA results (Hoyle and Duvall, 2004;S€ ur€ uc€ u et al, 2022). The multivariate normal distribution was tested using Mardia's test (Finch, 2013;Mardia, 1970).…”

Unveiling the Factors of Readiness for Change in the Digital Sharing Economy: An Exploratory Factor Analysis

“…Factor analysis is a statistical technique that can be employed to ascertain the underlying theoretical constructs that are present within a specific dataset, as well as to evaluate the degree to which these constructs accurately represent the original variables (Surucu et al, 2022). Moreover, factor analysis is a statistical technique that can be employed to examine the associations among observed data and to represent these associations using one or more latent variables.…”

Development and validation of a novel questionnaire regarding vision screening among preschool teachers in Malaysia

“…Eigenvalue more than 1, scree plot and cumulative percent of variance extracted were used as the criteria to determine the number of factors to remain [25][26] . In order to achieve a simple structure, each factor was made sure to have at least three variables and each variable must load only onto 1 factor [27] .…”

“…The removal and rearrangement of the items in attitude domain increased Cronbach's alpha value to 0.758. EFA is a statistical method to aid in making decision on retaining or removing an item in a questionnaire, apart from judgement by the research team to make sure the questionnaire has an accurate meaning to gather informative data for analysis [27] .…”