Accounting for Differential Item Functioning Using Bayesian Approximate Measurement Invariance

Sideridis, Georgios D.; Tsaousis, Ioannis; Alamri, Abeer A.

doi:10.1177/0013164419887482

Cited by 4 publications

(3 citation statements)

References 102 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A generalized version was implemented in Mplus by Asparouhov and Muthén (2017). The pppp-value can be used to evaluate the small variance parameters as demonstrated, for example, by Sideridis et al (2020). As Asparouhov and Muthén (2017) point out: If the pppp-value does not reject the model, it means that there is no evidence in the data for the parameters in the model to be outside the prior distribution.…”

Section: Bayesian Approximate Measurement Invariancementioning

confidence: 99%

Measurement invariance in the social sciences: Historical development, methodological challenges, state of the art, and future perspectives

Leitgöb¹,

Seddig²,

Asparouhov³

et al. 2023

Social Science Research

View full text Add to dashboard Cite

Section: Bayesian Approximate Measurement Invariancementioning

confidence: 99%

Measurement invariance in the social sciences: Historical development, methodological challenges, state of the art, and future perspectives

Leitgöb¹,

Seddig²,

Asparouhov³

et al. 2023

Social Science Research

View full text Add to dashboard Cite

“…Because full measurement invariance, which would also include equivalence of intercepts/thresholds (scalar invariance) and residuals (residual invariance) is unrealistic in large‐scale multicultural research (e.g. Davidov, Muthén, & Schmidt, 2018; Sideridis, Tsaousis, & Alamri, 2020), these results indicate that we can be reasonably confident that the YSR measures similar empirical constructs across diverse societies.…”

Section: Methodsmentioning

confidence: 92%

Effects of individual differences, society, and culture on youth‐rated problems and strengths in 38 societies

Ivanova

Achenbach

Turner

et al. 2022

Child Psychology Psychiatry

View full text Add to dashboard Cite

Background Clinicians increasingly serve youths from societal/cultural backgrounds different from their own. This raises questions about how to interpret what such youths report. Rescorla et al. (2019, European Child & Adolescent Psychiatry, 28, 1107) found that much more variance in 72,493 parents’ ratings of their offspring’s mental health problems was accounted for by individual differences than by societal or cultural differences. Although parents’ reports are essential for clinical assessment of their offspring, they reflect parents’ perceptions of the offspring. Consequently, clinical assessment also requires self‐reports from the offspring themselves. To test effects of individual differences, society, and culture on youths’ self‐ratings of their problems and strengths, we analyzed Youth Self‐Report (YSR) scores for 39,849 11–17 year olds in 38 societies. Methods Indigenous researchers obtained YSR self‐ratings from population samples of youths in 38 societies representing 10 culture cluster identified in the Global Leadership and Organizational Behavioral Effectiveness study. Hierarchical linear modeling of scores on 17 problem scales and one strengths scale estimated the percent of variance accounted for by individual differences (including measurement error), society, and culture cluster. ANOVAs tested age and gender effects. Results Averaged across the 17 problem scales, individual differences accounted for 92.5% of variance, societal differences 6.0%, and cultural differences 1.5%. For strengths, individual differences accounted for 83.4% of variance, societal differences 10.1%, and cultural differences 6.5%. Age and gender had very small effects. Conclusions Like parents’ ratings, youths’ self‐ratings of problems were affected much more by individual differences than societal/cultural differences. Most variance in self‐rated strengths also reflected individual differences, but societal/cultural effects were larger than for problems, suggesting greater influence of social desirability. The clinical significance of individual differences in youths’ self‐reports should thus not be minimized by societal/cultural differences, which—while important—can be taken into account with appropriate norms, as can gender and age differences.

show abstract

“…In practice, the full invariance or the partial invariance assumption are often only approximately fulfilled, and diversity of statistical methods has been proposed to tackle this case [32][33][34][35][36][37][38]. These approaches are of particular importance in studies of cross-cultural comparisons in which many groups (i.e., countries in this case) are involved [3,39].…”

Section: Full Invariance Partial Invariance and Linking Methodsmentioning

confidence: 99%

Lp Loss Functions in Invariance Alignment and Haberman Linking with Few or Many Groups

Robitzsch

2020

Stats

View full text Add to dashboard Cite

The comparison of group means in latent variable models plays a vital role in empirical research in the social sciences. The present article discusses an extension of invariance alignment and Haberman linking by choosing the robust power loss function ρ(x)=|x|p(p>0). This power loss function with power values p smaller than one is particularly suited for item responses that are generated under partial invariance. For a general class of linking functions, asymptotic normality of estimates is shown. Moreover, the theory of M-estimation is applied for obtaining linking errors (i.e., inference with respect to a population of items) for this class of linking functions. In a simulation study, it is shown that invariance alignment and Haberman linking have comparable performance, and in some conditions, the newly proposed robust Haberman linking outperforms invariance alignment. In three examples, the influence of the choice of a particular linking function on the estimation of group means is demonstrated. It is concluded that the choice of the loss function in linking is related to structural assumptions about the pattern of noninvariance in item parameters.

show abstract

Accounting for Differential Item Functioning Using Bayesian Approximate Measurement Invariance

Cited by 4 publications

References 102 publications

Measurement invariance in the social sciences: Historical development, methodological challenges, state of the art, and future perspectives

Measurement invariance in the social sciences: Historical development, methodological challenges, state of the art, and future perspectives

Effects of individual differences, society, and culture on youth‐rated problems and strengths in 38 societies

Lp Loss Functions in Invariance Alignment and Haberman Linking with Few or Many Groups

Contact Info

Product

Resources

About