Reinoud D. Stoel

Development of adolescent reading comprehension in Language 1 and Language 2: a longitudinal analysis of constituent components

This study investigated the relationship between reading comprehension development of 389 adolescents in their dominant language (Language 1 [L1], Dutch) and a foreign language (Language 2 [L2], English). In each consecutive year from Grades 8 through 10, a number of measurements were taken. Students reading comprehension, their linguistic knowledge (vocabulary and grammar knowledge) and processing efficiency (speed of word recognition and sentence comprehension) in both languages, and their metacognitive knowledge about reading were assessed. The relative strengths of the effects of these components of reading were analyzed to distinguish among 3 hypotheses about the relationship between L1 and L2 reading comprehension: the transfer hypothesis, the threshold hypothesis, and the processing efficiency hypothesis. The transfer hypothesis predicts a strong relationship between L1 and L2 reading comprehension and a strong effect of metacognitive knowledge on L2 reading comprehension, whereas the threshold and processing efficiency hypotheses predict a more important role of language-specific knowledge and processing skills. Results support the transfer hypothesis, although language-specific knowledge and fluency also contribute to L2 reading performance.

On the likelihood ratio test in structural equation modeling when parameters are subject to boundary constraints.

The authors show how the use of inequality constraints on parameters in structural equation models may affect the distribution of the likelihood ratio test. Inequality constraints are implicitly used in the testing of commonly applied structural equation models, such as the common factor model, the autoregressive model, and the latent growth curve model, although this is not commonly acknowledged. Such constraints are the result of the null hypothesis in which the parameter value or values are placed on the boundary of the parameter space. For instance, this occurs in testing whether the variance of a growth parameter is significantly different from 0. It is shown that in these cases, the asymptotic distribution of the chi-square difference cannot be treated as that of a central chi-square-distributed random variable with degrees of freedom equal to the number of constraints. The correct distribution for testing 1 or a few parameters at a time is inferred for the 3 structural equation models mentioned above. Subsequently, the authors describe and illustrate the steps that one should take to obtain this distribution. An important message is that using the correct distribution may lead to appreciably greater statistical power.

Predicting Growth Curves of Early Childhood Externalizing Problems: Differential Susceptibility of Children with Difficult Temperament

Using an accelerated longitudinal design, the development of externalizing problems from age 2 to 5xa0years was investigated in relation to maternal psychopathology, maternal parenting, gender, child temperament, and the presence of siblings. The sample consisted of 150 children selected at age 2–3xa0years for having high levels of externalizing problems. Parenting was measured using observational methods, and maternal reports were used for the other variables. Overall, mean levels of externalizing problems decreased over time, and higher initial levels (intercept) were related to a stronger decrease (negative slope) in externalizing problems. Results showed that higher levels of maternal psychopathology were related to less decrease in early childhood externalizing problems. Parental sensitive behavior predicted a stronger decrease in externalizing problems, but only for children with difficult temperaments. A stronger decrease of externalizing problems in children with older siblings also pertained only to children with difficult temperaments. Thus, temperamentally difficult children appear to be more susceptible to environmental influences on the development of externalizing behaviors. Our results indicate that the role of siblings in early childhood externalizing problems deserves more research attention, and that intervention efforts need to take into account temperamental differences in children’s susceptibility to environmental influences.

Testing Measurement Invariance in the Target Rotated Multigroup Exploratory Factor Model.

We propose a method to investigate measurement invariance in the multigroup exploratory factor model, subject to target rotation. We consider both oblique and orthogonal target rotation. This method has clear advantages over other approaches, such as the use of congruence measures. We demonstrate that the model can be implemented readily in the freely available Mx program. We present the results of 2 illustrative analyses, one based on artificial data, and the other on real data relating to personality in male and female psychology students.

A Note on Testing Perfect Correlations in SEM

This article is concerned with the seemingly simple problem of testing whether latent factors are perfectly correlated (i.e., statistically indistinct). In recent literature, researchers have used different approaches, which are not always correct or complete. We discuss the parameter constraints required to obtain such perfectly correlated latent factors in the context of 4 commonly used models: (a) the oblique factor model, (b) the hierarchical factor model, (c) models in which the factors are predicted by a covariate, and (d) models in which the factors are predictors of a dependent variable. It is shown that the necessary constraints depend on the choice of scaling. We illustrate testing the indistinctiveness of factors with 2 real data examples.

Methodological Issues in the Application of the Latent Growth Curve Model

Latent growth curve analysis (McArdle, 1986, 1988; Meredith & Tisak, 1990; Willett & Sayer, 1994) is well suited to analyze systematic change in longitudinal data collected from a panel design. It represents outcome variables explicitly as a function of time and other measures. Specifically, latent growth curve analysis is a Statistical technique to estimate the Parameters that represent the growth curves that are assumed to have given rise to the structure of the repeatedly measured outcome variable over time. Growth curve analysis can be applied just to get a (unconditional) description of the mean growth over a certain period of time. However, the emphasis of this technique lies in explanation of differences between subjects in the parameters describing the growth curves; in other words, in the systematic inter-individual differences in intra-individual change.

Efficiency of Visual Information Processing in Children at-Risk for Dyslexia: Habituation of Single-Trial ERPs.

To investigate underlying learning mechanisms in relation to the development of dyslexia, event-related potentials to visual standards were recorded in five-year-old pre-reading children at-risk for familial dyslexia (n=24) and their controls (n=14). At the end of second grade the children aged 8 years were regrouped into three groups according to literacy level and risk factor. Single-trial analyses revealed N1 habituation in the normal-reading controls, but not in the normal-reading at-risks, and a N1 amplitude increase in the group of poor-reading at-risks and poor-reading controls. No P3 habituation was found in either group. The normal-reading at-risk group exhibited the longest N1 and P3 latencies, possibly compensating for their reduced neuronal activity during initial information extraction. In contrast, the poor-reading group only showed prolonged P3, and their increase in (initial small) N1 amplitude together with normal N1 latencies, suggests inefficient processing in an early time window, which might explain automatisation difficulties in dyslexic readers.

Including Time-Invariant Covariates in the Latent Growth Curve Model

Within the latent growth curve model, time-invariant covariates are generally modeled on the subject level, thereby estimating the effect of the covariate on the latent growth parameters. Incorporating the time-invariant covariate in this manner may have some advantages regarding the interpretation of the effect but may also be incorrect in certain instances. In this article we discuss a more general approach for modeling time-invariant covariates in latent growth curve models in which the covariate is directly regressed on the observed indicators. The approach can be used on its own to get estimates of the growth curves corrected for the influence of a 3rd variable, or it can be used to test the appropriateness of the standard way of modeling the time-invariant covariates. It thus provides a test of the assumption of full mediation, which states that the relation between the covariate and the observed indicators is fully mediated by the latent growth parameters.

Teachers’ assessment of antisocial behavior in kindergarten : physical aggression and measurement bias across gender

A confirmatory factor analytic study was conducted to obtain evidence for physical aggression as a distinct construct of nonaggressive antisocial behavior in young children. Second, the authors investigated factorial invariance across gender. Teachers completed the Preschool Behavior Questionnaire (PBQ) for two independent samples of kindergartners (N = 487; N = 1,557). Behavior items were selected representing physically aggressive versus nonaggressive antisocial behavior. To obtain support for the two-factor model, the authors also examined associations with subtypes of internalizing behavior. Confirmatory factor analyses confirmed that physical aggression constitutes a distinct construct from nonaggressive antisocial behavior for young children. In support of the model, differential associations with internalizing behavior and different outcomes with respect to gender differences were found. Factorial noninvariance across gender was found for physical aggression, and explanations for those gender differences are discussed.

Teachers' Assessment of Physical Aggression with the Preschool Behavior Questionnaire: A Multitrait-Multimethod Evaluation of Convergent and Discriminant Validity.

The distinctiveness of physical aggression from other antisocial behavior is widely accepted but little research has explicitly focused on young children to empirically test this assumption. A Multitrait-Multimethod Matrix (MTMM) approach was employed to confirm the distinctiveness of physical aggression from nonaggressive antisocial behavior in early childhood. In addition, the convergent validity of teacher reports of physical aggression was investigated on a measure that contained age-appropriate behavior items selected from the Preschool Behavior Questionnaire (PBQ). Assessments of physical aggression versus nonaggressive antisocial behavior of 117 kindergartners with different behavior profiles were obtained using three measures varying in source and/or method: (a) teacher reports on the PBQ, (b) short interviews with teachers, and (c) classroom observations. The MTMM matrix was analyzed using confirmatory factor analysis (CFA). The results provided reasonable support for the discrimination between physical aggression and nonaggressive antisocial behavior. In addition, strong support was found for the convergent validity of teacher-reported physical aggression using PBQ items.