Ab Mooijaart

How work and family can facilitate each other: Distinct types of work-family facilitation and outcomes for women and men.

This study was designed to gain more insight in the different ways in which work and family roles can benefit each other. Both qualitative (N=25) and quantitative (N=352) results obtained in a financial service organization supported the distinction between energy-based, time-based, behavioral, and psychological work-family facilitation, in addition to different types of work-family conflict that were identified in previous research. As expected, facilitation contributed substantially and differentially to the prediction of work and nonwork outcomes, over and above the effects of conflict. As predicted, women experienced higher levels of facilitation than men did. Furthermore, results indicate that examining facilitation, in addition to conflict, is especially important to predict the work and home life experiences of women.

Factor analysis for non-normal variables

Factor analysis for nonnormally distributed variables is discussed in this paper. The main difference between our approach and more traditional approaches is that not only second order cross-products (like covariances) are utilized, but also higher order cross-products. It turns out that under some conditions the parameters (factor loadings) can be uniquely determined. Two estimation procedures will be discussed. One method gives Best Generalized Least Squares (BGLS) estimates, but is computationally very heavy, in particular for large data sets. The other method is a least squares method which is computationally less heavy. In one example the two methods will be compared by using the bootstrap method. In another example real life data are analyzed.

Cannabis use and development of externalizing and internalizing behaviour problems in early adolescence: A TRAILS study

AIM To examine the prospective relationship between externalizing and internalizing problems and cannabis use in early adolescence. MATERIALS AND METHODS Data were used from the TRAILS study, a longitudinal cohort study of (pre)adolescents (n=1,449), with measurements at age 11.1 (T1), age 13.6 (T2) and age 16.3 (T3). Internalizing (withdrawn behaviour, somatic complaints and depression) and externalizing (delinquent and aggressive behaviour) problems were assessed at all data waves, using the Youth Self Report. Participants reported on cannabis use at the second and third wave. Path analysis was used to identify the temporal order of internalizing and externalizing problems and cannabis use. RESULTS Path analysis showed no associations between cannabis use (T2-T3) and internalizing problems (T1-2-3). However, cannabis use and externalizing problems were associated (r ranged from .19-.58); path analysis showed that externalizing problems at T1 and T2 preceded cannabis use at T2 and T3, respectively. In contrast, cannabis use (T2) did not precede externalizing problems (T3). CONCLUSIONS These results suggest that in early adolescence, there is no association between internalizing behaviour and cannabis use. There is an association between externalizing behaviour and cannabis use, and it appears that externalizing behaviour precedes cannabis use rather than the other way around during this age period.

Type I errors and power of the parametric bootstrap goodness‐of‐fit test: Full and limited information

In sparse tables for categorical data well-known goodness-of-fit statistics are not chi-square distributed. A consequence is that model selection becomes a problem. It has been suggested that a way out of this problem is the use of the parametric bootstrap. In this paper, the parametric bootstrap goodness-of-fit test is studied by means of an extensive simulation study; the Type I error rates and power of this test are studied under several conditions of sparseness. In the presence of sparseness, models were used that were likely to violate the regularity conditions. Besides bootstrapping the goodness-of-fit usually used (full information statistics), corrected versions of these statistics and a limited information statistic are bootstrapped. These bootstrap tests were also compared to an asymptotic test using limited information. Results indicate that bootstrapping the usual statistics fails because these tests are too liberal, and that bootstrapping or asymptotically testing the limited information statistic works better with respect to Type I error and outperforms the other statistics by far in terms of statistical power. The properties of all tests are illustrated using categorical Markov models.

An Alternative Approach for Nonlinear Latent Variable Models

In the last decades there has been an increasing interest in nonlinear latent variable models. Since the seminal paper of Kenny and Judd, several methods have been proposed for dealing with these kinds of models. This article introduces an alternative approach. The methodology involves fitting some third-order moments in addition to the means and covariances. This article discusses how the model equations can be formulated and how several standard tests, like the model fit and Lagrange multiplier tests, can be performed. The new method compares favorably with the maximum likelihood method in several studies and can provide evidence of interaction that earlier approaches might ignore.

The EM algorithm for latent class analysis with equality constraints

The EM algorithm is a popular iterative method for estimating parameters in the latent class model where at each step the unknown parameters can be estimated simply as weighted sums of some latent proportions. The algorithm may also be used when some parameters are constrained to equal given constants or each other. It is shown that in the general case with equality constraints, the EM algorithm is not simple to apply because a nonlinear equation has to be solved. This problem arises, mainly, when equality constrints are defined over probabilities indifferent combinations of variables and latent classes. A simple condition is given in which, although probabilities in different variable-latent class combinations are constrained to be equal, the EM algorithm is still simple to apply.

A general solution of the weighted orthonormal procrustes problem

A general solution for weighted orthonormal Procrustes problem is offered in terms of the least squares criterion. For the two-demensional case. this solution always gives the global minimum; for the general case, an algorithm is proposed that must converge, although not necessarily to the global minimum. In general, the algorithm yields a solution for the problem of how to fit one matrix to another under the condition that the dimensions of the latter matrix first are allowed to be transformed orthonormally and then weighted differentially, which is the task encountered in fitting analogues of the IDIOSCAL and INDSCAL models to a set of configurations.

Some New Log-Bilinear Models for the Analysis of Asymmetry in a Square Contingency Table

The authors propose some new log-bilinear models for the analysis of asymmetry in square contingency tables. In these models the logarithm of the expected frequency is split up into two parts: (a) a symmetric or a quasi-symmetric part, and (b) a skew-symmetric part. The skew-symmetric part is decomposed using the so-called Gower decomposition of skew-symmetric matrices. It is shown how graphical representations of this decomposition can simplify the interpretation.

Three-factor association models for three-way contingency tables

Abstract A general formulation of association models is introduced for the analysis of three-way contingency tables. The two-factor and three-factor interaction matrices are decomposed into matrices of lower rank. In particular, the three-factor interaction is decomposed by the PARAFAC model. Various restricted models can be used to validate special assumptions for the data such as departures from conditional independence in the context of sets of contingency tables. The problem of identification is discussed. Two sets of data are analyzed to illustrate the versatility in the interpretation and the advantages of the models and methods developed here.

Estimating the Statistical Power in Small Samples by Empirical Distributions

Model selection is an important issue in Structural Equation Modeling. This issue is in particular important if the sample size is small. In such a case the power may be very small, so it is hard to decide which model is the most appropriate, because alternative models may also fit the data well. There is a lot of theoretical papers on this subject, however they are almost always dealing with asymptotic theory. In cases with small samples this assumption may lead to wrong results. We will use resampling methods, like the parametric bootstrap, to investigate the empirical distribution of certain statistics. On the basis of this empirical distribution we can investigate the power of some tests, even in cases with small samples. An example will be given.