Estimating statistical power for structural equation models in developmental cognitive science: A tutorial in R : Power simulation for SEMs.

Determining the compositional structure and dimensionality of psychological constructs lies at the heart of many research questions in developmental science. Structural equation modeling (SEM) provides a versatile framework for formalizing and estimating the relationships among multiple latent constructs. While the flexibility of SEM can accommodate many complex assumptions on the underlying structure of psychological constructs, it makes a priori estimation of statistical power and required sample size challenging. This difficulty is magnified when comparing non-nested SEMs, which prevents the use of traditional likelihood-ratio tests. Sample size estimates for SEM model fit comparisons typically rely on generic rules of thumb. Such heuristics can be misleading because statistical power in SEM depends on a variety of model properties. Here, we demonstrate a Monte Carlo simulation approach for estimating a priori statistical power for model selection when comparing non-nested models in an SEM framework. We provide a step-by-step guide to this approach based on an example from our memory development research in children.

A Note on the Connection Between Trek Rules and Separable Nonlinear Least Squares in Linear Structural Equation Models.

We show that separable nonlinear least squares (SNLLS) estimation is applicable to all linear structural equation models (SEMs) that can be specified in RAM notation. SNLLS is an estimation technique that has successfully been applied to a wide range of models, for example neural networks and dynamic systems, often leading to improvements in convergence and computation time. It is applicable to models of a special form, where a subset of parameters enters the objective linearly. Recently, Kreiberg et al. (Struct Equ Model Multidiscip J 28(5):725-739, 2021. https://doi.org/10.1080/10705511.2020.1835484 ) have shown that this is also the case for factor analysis models. We generalize this result to all linear SEMs. To that end, we show that undirected effects (variances and covariances) and mean parameters enter the objective linearly, and therefore, in the least squares estimation of structural equation models, only the directed effects have to be obtained iteratively. For model classes without unknown directed effects, SNLLS can be used to analytically compute least squares estimates. To provide deeper insight into the nature of this result, we employ trek rules that link graphical representations of structural equation models to their covariance parametrization. We further give an efficient expression for the gradient, which is crucial to make a fast implementation possible. Results from our simulation indicate that SNLLS leads to improved convergence rates and a reduced number of iterations.

Eleven strategies for making reproducible research and open science training the norm at research institutions.

Reproducible research and open science practices have the potential to accelerate scientific progress by allowing others to reuse research outputs, and by promoting rigorous research that is more likely to yield trustworthy results. However, these practices are uncommon in many fields, so there is a clear need for training that helps and encourages researchers to integrate reproducible research and open science practices into their daily work. Here, we outline eleven strategies for making training in these practices the norm at research institutions. The strategies, which emerged from a virtual brainstorming event organized in collaboration with the German Reproducibility Network, are concentrated in three areas: (i) adapting research assessment criteria and program requirements; (ii) training; (iii) building communities. We provide a brief overview of each strategy, offer tips for implementation, and provide links to resources. We also highlight the importance of allocating resources and monitoring impact. Our goal is to encourage researchers - in their roles as scientists, supervisors, mentors, instructors, and members of curriculum, hiring or evaluation committees - to think creatively about the many ways they can promote reproducible research and open science practices in their institutions.

How Specific Abilities Might Throw 'g' a Curve: An Idea on How to Capitalize on the Predictive Validity of Specific Cognitive Abilities.

School grades are still used by universities and employers for selection purposes. Thus, identifying determinants of school grades is important. Broadly, two predictor categories can be differentiated from an individual difference perspective: cognitive abilities and personality traits. Over time, evidence accumulated supporting the notion of the g-factor as the best single predictor of school grades. Specific abilities were shown to add little incremental validity. The current paper aims at reviving research on which cognitive abilities predict performance. Based on ideas of criterion contamination and deficiency as well as Spearman's ability differentiation hypothesis, two mechanisms are suggested which both would lead to curvilinear relations between specific abilities and grades. While the data set provided for this special issue does not allow testing these mechanisms directly, we tested the idea of curvilinear relations. In particular, polynomial regressions were used. Machine learning was applied to identify the best fitting models in each of the subjects math, German, and English. In particular, we fitted polynomial models with varying degrees and evaluated their accuracy with a leave-one-out validation approach. The results show that tests of specific abilities slightly outperform the g-factor when curvilinearity is assumed. Possible theoretical explanations are discussed.