How to factor-analyze your data right: do’s, don’ts, and how-to’s.

By Masaki Matsunaga. 2010

1,426 citations recorded by the source corpus.

The current article provides a guideline for conducting factor analysis, a technique used to estimate the population-level factor structure underlying the given sample data. First, the distinction between exploratory and confirmatory factor analyses (EFA and CFA) is briefly discussed; along with this discussion, the notion of principal component analysis and why it does not provide a valid substitute of factor analysis is noted. Second, a step-by-step walk-through of conducting factor analysis is illustrated; through these walk-through instructions, various decisions that need to be made in factor analysis are discussed and recommendations provided. Specifically, suggestions for how to carry out preliminary procedures, EFA, and CFA are provided with SPSS and LISREL syntax examples. Finally, some critical issues concerning the appropriate (and not-so-appropriate) use of factor analysis are discussed along with the discussion of recommended practices.

Subfields: Computer Networks and Communications, Management Science and Operations Research, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/1535fea223fabfe05a226b2d76a4b40524710f97

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

In Management Science and Operations Research

In Statistics and Probability

Citations in the corpus

Establishing the theoretical components of alexithymia via factor analysis: Introduction and validation of the attention-appraisal model of alexithymia (2017, methodology): “Factor loadings > .40 were considered meaningful loadings; this cut-off was selected because it has been endorsed by various authors (e.g., Matsunaga, 2010; Stevens, 1992) and is the most commonly used cut-off within psychometric literature (e.g., Bagby et al., 2006; Kooiman et al., 2002).”
One Size Doesn’t Fit All: Using Factor Analysis to Gather Validity Evidence When Using Surveys in Your Research (2019, background): “30 are usually considered to be problematic (Matsunaga, 2010), especially if the item does not have a particularly strong loading on a focal factor.”
The Development and Validation of the Game User Experience Satisfaction Scale (GUESS) (2016, methodology): “In terms of the cutoff value for item loading, the |0.40| value was selected because it is the most common value and it falls in the recommend range of cutoff values (Hinkin, 1995, 1998; Field, 2009; Matsunaga, 2010; Nunnally, 1978; Tabachnick & Fidell, 2007).”
Passive and Active Social Media Use and Depressive Symptoms Among United States Adults (2018, neutral): “4).(52) Thus, this item was removed from the analysis.”
Parenting Interactions with Children: Checklist of Observations Linked to Outcomes (PICCOLO) in Diverse Ethnic Groups (2013, background): “40 (Matsunaga, 2010), with an overall average factor loading of .”
Measuring Tourists’ Emotional Solidarity with One Another—A Modification of the Emotional Solidarity Scale (2020, background): “The 0.55 cutoff is equal to what Comrey and Lee (1992) considered good loading and is a rather conservative cutoff (Matsunaga 2010).”
U.S. teachers' conceptions of the purposes of assessment (2017, methodology): “Furthermore, because the promax method retains a solution with the lowest kappa, it leads to factor structures that are “maximally distinguishable” (Matsunaga, 2010).”
Measuring Teacher Engagement: Development of the Engaged Teachers Scale (ETS) (2013, methodology): “Although the use of PCA has been criticised as a means of extracting factors (e.g., Velicer & Jackson, 1990), it is a preferred method for item reduction (Conway & Huffcutt, 2003; Henson & Roberts, 2006; Matsunaga, 2010).”
Does the Possibility to Make Equity Investments in Crowdfunding Projects Crowd Out Reward–Based Investments? (2015, methodology): “In our case, given that the secondary loading of 0.42 is higher than the keep threshold, we exclude this item from the financial motivation subscale (see Matsunaga, 2010).”