Assessing measurement invariance with moderated nonlinear factor analysis using the R package OpenMx.

Kolbe, Laura; Molenaar, Dylan; Jak, Suzanne; Jorgensen, Terrence D.

doi:10.1037/met0000501

Cited by 12 publications

(15 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Recalling that an arrow pointing directly at the factor represents a shift in the factor mean, we can see that, in contrast, there is mean impact associated with all of the background variables. Example M plus syntax for fitting a MNLFA model with the pattern of impact and DIF displayed in Figure 6 is provided in Appendix S6 (see also Kolbe et al, 2022, for a description of how to fit MNLFA models with the R package OpenMx).…”

Section: Mnlfamentioning

confidence: 99%

“…One option, pursued for MNLFA in recent work by Bauer et al (2020), Chen et al (2022), and Brandt et al (2023), is to eschew traditional inferential approaches to DIF detection (e.g. likelihood ratio tests, Wald tests, or Lagrange multipliers/modification indices/score tests) and instead use regularized estimation to automate the process.…”

Section: Practical Considerationsmentioning

confidence: 99%

“…DIF selection can thus be achieved in an automated way that avoids the use of inference tests. Within a Bayesian setting, regularizing priors on the DIF parameters, such as the spike and slab, take on a role similar to the penalty (Brandt et al, 2023; Chen et al, 2022). Again, the idea is to obtain a sparse‐DIF solution without recourse to a multitude of inference tests.…”

Section: Practical Considerationsmentioning

confidence: 99%

See 2 more Smart Citations

Enhancing measurement validity in diverse populations: Modern approaches to evaluating differential item functioning

Bauer

2023

Brit J Math & Statis

View full text Add to dashboard Cite

When developing and evaluating psychometric measures, a key concern is to ensure that they accurately capture individual differences on the intended construct across the entire population of interest. Inaccurate assessments of individual differences can occur when responses to some items reflect not only the intended construct but also construct‐irrelevant characteristics, like a person's race or sex. Unaccounted for, this item bias can lead to apparent differences on the scores that do not reflect true differences, invalidating comparisons between people with different backgrounds. Accordingly, empirically identifying which items manifest bias through the evaluation of differential item functioning (DIF) has been a longstanding focus of much psychometric research. The majority of this work has focused on evaluating DIF across two (or a few) groups. Modern conceptualizations of identity, however, emphasize its multi‐determined and intersectional nature, with some aspects better represented as dimensional than categorical. Fortunately, many model‐based approaches to modelling DIF now exist that allow for simultaneous evaluation of multiple background variables, including both continuous and categorical variables, and potential interactions among background variables. This paper provides a comparative, integrative review of these new approaches to modelling DIF and clarifies both the opportunities and challenges associated with their application in psychometric research.

show abstract

Section: Mnlfamentioning

confidence: 99%

Section: Practical Considerationsmentioning

confidence: 99%

Section: Practical Considerationsmentioning

confidence: 99%

See 1 more Smart Citation

Enhancing measurement validity in diverse populations: Modern approaches to evaluating differential item functioning

Bauer

2023

Brit J Math & Statis

View full text Add to dashboard Cite

show abstract

“…Nevertheless, the examination of gender was based on a categorical variable (gender binary). Recent advances in measurement invariance analysis have enabled the test of invariance within a dimensional perspective [40]. Thus, further studies should investigate the invariance of the BDI-II across gender dimensions that may allow for more inclusive conclusions.…”

Section: Limitations and Futurementioning

confidence: 99%

Comprehensive investigation of factor structure and gender equivalence of the Beck Depression Inventory-II among nonclinical adolescents

Ferreira-Maia

Gorenstein

Wang

2023

Preprint

View full text Add to dashboard Cite

The Beck Depression Inventory-II (BDI-II) is a tool widely used to rate the severity of depressive symptoms. Regarding the factor validity of the BDI-II in adolescents, the two-factor solution describing “cognitive” and “somatic-affective” dimensions has been shown to be the most reported. Recently, bifactor model analysis has become popular to describe psychological constructs such as depression. Furthermore, data related to measurement invariance across gender is limited. This study aimed to examine plausible factor structures and gender equivalence of the BDI-II in a Brazilian sample of nonclinical adolescents (N = 1,184, aged 13–18 years, 59.1% females). Cross-validation of the BDI-II was performed through exploratory (EFA) and confirmatory factor analysis (CFA). Measurement invariance was evaluated using multigroup-CFA (MG-CFA). EFA suggested an oblique two-factor model depicting “affective-cognitive” and “somatic" dimensions. Simple structure and bifactor models from literature, as well as the estimated EFA model and its corresponding bifactor model were examined. All factor models revealed adequate and similar fitness, well-defined factors, and good reliability. The goodness-of-fit of bifactor models was slightly superior when compared to simple structure models. Bifactor analyses indicated a robust general factor with low reliable variance in total scores attributed to multidimensionality caused by the group factors in bifactor models. MG-CFA supported invariance across gender, confirming that the same BDI-II construct could be applied to both female and male adolescents. This study provides evidence that the BDI-II could be used by researchers and clinicians as a unidimensional measure of depressive symptoms in the adolescent population.

show abstract

“…MNLFA can be estimated in R (R Core Team, 2021) via the package OpenMx (Boker et al, 2011). We refer readers to Kolbe et al (2022) for a detailed guide on how to estimate MNLFA in OpenMx and specifically how to use it to investigate MI.…”

Section: A More Holistic View On Measurement Invariancementioning

confidence: 99%

A Causal Framework for the Comparability of Latent Variables

Sterner,

Pargent,

Deffner

et al. 2023

Preprint

View full text Add to dashboard Cite

Measurement invariance (MI) describes the equivalence of measurement models of a construct across groups or time. When comparing latent means, MI is often stated as a prerequisite of meaningful group comparisons. The most common way to investigate MI is multi-group confirmatory factor analysis (MG-CFA). Although numerous guides exist, a recent review showed that MI is rarely investigated in practice. We argue that one reason might be that the results of MG-CFA are uninformative as to why MI does not hold between groups. Consequently, under this framework, it is difficult to regard the study of MI an interesting and constructive step in the modeling process. We show how directed acyclic graphs (DAGs) from the causal inference literature can guide researchers in reasoning about the causes of non-invariance. For this, we first show how DAGs for measurement models can be translated into the path diagrams used in the linear structural equation model (SEM) literature. We then demonstrate how insights gained from this causal perspective can be used to explicitly model encoded causal assumptions with moderated SEMs, allowing for a more enlightening investigation of MI. Ultimately, our goal is to provide a framework in which the investigation of MI is not deemed a “gateway test” that simply licenses further analyses. By enabling researchers to consider MI as an interesting part of the modeling process, we hope to increase the prevalence of investigations of MI altogether.

show abstract

Assessing measurement invariance with moderated nonlinear factor analysis using the R package OpenMx.

Cited by 12 publications

References 47 publications

Enhancing measurement validity in diverse populations: Modern approaches to evaluating differential item functioning

Enhancing measurement validity in diverse populations: Modern approaches to evaluating differential item functioning

Comprehensive investigation of factor structure and gender equivalence of the Beck Depression Inventory-II among nonclinical adolescents

A Causal Framework for the Comparability of Latent Variables

Contact Info

Product

Resources

About