Mariska Barendse

Publications

Displaying 1 - 10 of 10
  • Barendse, M. T., & Rosseel, Y. (2020). Multilevel modeling in the ‘wide format’ approach with discrete data: A solution for small cluster sizes. Structural Equation Modeling: A Multidisciplinary Journal, 27(5), 696-721. doi:10.1080/10705511.2019.1689366.

    Abstract

    In multilevel data, units at level 1 are nested in clusters at level 2, which in turn may be nested in even larger clusters at level 3, and so on. For continuous data, several authors have shown how to model multilevel data in a ‘wide’ or ‘multivariate’ format approach. We provide a general framework to analyze random intercept multilevel SEM in the ‘wide format’ (WF) and extend this approach for discrete data. In a simulation study, we vary response scale (binary, four response options), covariate presence (no, between-level, within-level), design (balanced, unbalanced), model misspecification (present, not present), and the number of clusters (small, large) to determine accuracy and efficiency of the estimated model parameters. With a small number of observations in a cluster, results indicate that the WF approach is a preferable approach to estimate multilevel data with discrete response options.
  • Burgers, N., Ettema, D. F., Hooimeijer, P., & Barendse, M. T. (2020). The effects of neighbours on sport club membership. European Journal for Sport and Society. Advance online publication. doi:10.1080/16138171.2020.1840710.

    Abstract

    Neighbours have been found to influence each other’s behaviour (contagion effect). However, little is known about the influence on sport club membership. This while increasing interest has risen for the social role of sport clubs. Sport clubs could bring people from different backgrounds together. A mixed composition is a key element in this social role. Individual characteristics are strong predictors of sport club membership. Western high educated men are more likely to be members. In contrast to people with a non-Western migration background. The neighbourhood is a more fixed meeting place, which provides unique opportunities for people from different backgrounds to interact. This study aims to gain more insight into the influence of neighbours on sport club membership. This research looks especially at the composition of neighbour’s migration background, since they tend to be more or less likely to be members and therefore could encourage of inhibit each other. A population database including the only registry data of all Dutch inhabitants was merged with data of 11 sport unions. The results show a cross-level effect of neighbours on sport club membership. We find a contagion effect of neighbours’ migration background; having a larger proportion of neighbours with a migration background from a non-Western country reduces the odds, as expected. However, this contagion effect was not found for people with a Moroccan or Turkish background.
  • De Smedt, F., Merchie, E., Barendse, M. T., Rosseel, Y., De Naeghel, J., & Van Keer, H. (2018). Cognitive and motivational challenges in writing: Studying the relation with writing performance across students' gender and achievement level. Reading Research Quarterly, 53(2), 249-272. doi:10.1002/rrq.193.

    Abstract

    Abstract In the past, several assessment reports on writing repeatedly showed that elementary school students do not develop the essential writing skills to be successful in school. In this respect, prior research has pointed to the fact that cognitive and motivational challenges are at the root of the rather basic level of elementary students' writing performance. Additionally, previous research has revealed gender and achievement-level differences in elementary students' writing. In view of providing effective writing instruction for all students to overcome writing difficulties, the present study provides more in-depth insight into (a) how cognitive and motivational challenges mediate and correlate with students' writing performance and (b) whether and how these relations vary for boys and girls and for writers of different achievement levels. In the present study, 1,577 fifth- and sixth-grade students completed questionnaires regarding their writing self-efficacy, writing motivation, and writing strategies. In addition, half of the students completed two writing tests, respectively focusing on the informational or narrative text genre. Based on multiple group structural equation modeling (MG-SEM), we put forward two models: a MG-SEM model for boys and girls and a MG-SEM model for low, average, and high achievers. The results underline the importance of studying writing models for different groups of students in order to gain more refined insight into the complex interplay between motivational and cognitive challenges related to students' writing performance.
  • Barendse, M. T., Ligtvoet, R., Timmerman, M. E., & Oort, F. J. (2016). Model fit after pairwise maximum likelihood. Frontiers in Psychology, 7: 528. doi:10.3389/fpsyg.2016.00528.

    Abstract

    Maximum likelihood factor analysis of discrete data within the structural equation modeling framework rests on the assumption that the observed discrete responses are manifestations of underlying continuous scores that are normally distributed. As maximizing the likelihood of multivariate response patterns is computationally very intensive, the sum of the log–likelihoods of the bivariate response patterns is maximized instead. Little is yet known about how to assess model fit when the analysis is based on such a pairwise maximum likelihood (PML) of two–way contingency tables. We propose new fit criteria for the PML method and conduct a simulation study to evaluate their performance in model selection. With large sample sizes (500 or more), PML performs as well the robust weighted least squares analysis of polychoric correlations.
  • Barendse, M. T. (2015). Dimensionality assessment with factor analysis methods. PhD Thesis, University of Groningen, Groningen.
  • Barendse, M. T., Oort, F. J., & Timmerman, M. E. (2015). Using exploratory factor analysis to determine the dimensionality of discrete responses. Structural Equation Modeling: A Multidisciplinary Journal, 22(1), 87-101. doi:10.1080/10705511.2014.934850.

    Abstract

    Exploratory factor analysis (EFA) is commonly used to determine the dimensionality of continuous data. In a simulation study we investigate its usefulness with discrete data. We vary response scales (continuous, dichotomous, polytomous), factor loadings (medium, high), sample size (small, large), and factor structure (simple, complex). For each condition, we generate 1,000 data sets and apply EFA with 5 estimation methods (maximum likelihood [ML] of covariances, ML of polychoric correlations, robust ML, weighted least squares [WLS], and robust WLS) and 3 fit criteria (chi-square test, root mean square error of approximation, and root mean square residual). The various EFA procedures recover more factors when sample size is large, factor loadings are high, factor structure is simple, and response scales have more options. Robust WLS of polychoric correlations is the preferred method, as it is theoretically justified and shows fewer convergence problems than the other estimation methods.
  • Barendse, M. T., Albers, C. J., Oort, F. J., & Timmerman, M. E. (2014). Measurement bias detection through Bayesian factor analysis. Frontiers in Psychology, 5: 1087. doi:10.3389/fpsyg.2014.01087.

    Abstract

    Measurement bias has been defined as a violation of measurement invariance. Potential violators—variables that possibly violate measurement invariance—can be investigated through restricted factor analysis (RFA). The purpose of the present paper is to investigate a Bayesian approach to estimate RFA models with interaction effects, in order to detect uniform and nonuniform measurement bias. Because modeling nonuniform bias requires an interaction term, it is more complicated than modeling uniform bias. The Bayesian approach seems especially suited for such complex models. In a simulation study we vary the type of bias (uniform, nonuniform), the type of violator (observed continuous, observed dichotomous, latent continuous), and the correlation between the trait and the violator (0.0, 0.5). For each condition, 100 sets of data are generated and analyzed. We examine the accuracy of the parameter estimates and the performance of two bias detection procedures, based on the DIC fit statistic, in Bayesian RFA. Results show that the accuracy of the estimated parameters is satisfactory. Bias detection rates are high in all conditions with an observed violator, and still satisfactory in all other conditions.
  • Barendse, M. T., Oort, F. J., Jak, S., & Timmerman, M. E. (2013). Multilevel exploratory factor analysis of discrete data. Netherlands Journal of Psychology, 67(4), 114-121.
  • Barendse, M. T., Oort, F. J., Werner, C. S., Ligtvoet, R., & Schermelleh-Engel, K. (2012). Measurement bias detection through factor analysis. Structural Equation Modeling: A Multidisciplinary Journal, 19(4), 561-579. doi:10.1080/10705511.2012.713261.

    Abstract

    Measurement bias is defined as a violation of measurement invariance, which can be investigated through multigroup factor analysis (MGFA), by testing across-group differences in intercepts (uniform bias) and factor loadings (nonuniform bias). Restricted factor analysis (RFA) can also be used to detect measurement bias. To also enable nonuniform bias detection, we extend RFA with latent moderated structures (LMS) or use a random slope parameterization (RSP). In a simulation study we compare the MGFA, RFA/LMS, and RFA/RSP methods in detecting measurement bias, varying type of bias (uniform, nonuniform), type of the variable that violates measurement invariance (dichotomous, continuous), and its relationship with the trait that we want to measure (independent, dependent). For each condition, 500 sets of data are generated and analyzed with each of the three detection methods, in single run and in an iterative procedure. The RFA methods outperform MGFA when the violating variable is continuous.
  • Barendse, M. T., Oort, F. J., & Garst, G. J. A. (2010). Using restricted factor analysis with latent moderated structures to detect uniform and nonuniform measurement bias: A simulation study. AStA Advances in Statistical Analysis, 94, 117-127. doi:10.1007/s10182-010-0126-1.

    Abstract

    Factor analysis is an established technique for the detection of measurement bias. Multigroup factor analysis (MGFA) can detect both uniform and nonuniform bias. Restricted factor analysis (RFA) can also be used to detect measurement bias, albeit only uniform measurement bias. Latent moderated structural equations (LMS) enable the estimation of nonlinear interaction effects in structural equation modelling. By extending the RFA method with LMS, the RFA method should be suited to detect nonuniform bias as well as uniform bias. In a simulation study, the RFA/LMS method and the MGFA method are compared in detecting uniform and nonuniform measurement bias under various conditions, varying the size of uniform bias, the size of nonuniform bias, the sample size, and the ability distribution. For each condition, 100 sets of data were generated and analysed through both detection methods. The RFA/LMS and MGFA methods turned out to perform equally well. Percentages of correctly identified items as biased (true positives) generally varied between 92% and 100%, except in small sample size conditions in which the bias was nonuniform and small. For both methods, the percentages of false positives were generally higher than the nominal levels of significance.

Share this page