![]() ![]() Although researchers in psychology commonly encounter variable selection problems, many researchers remain unaware of suitable variable selection methods.Īs noted by McNeish ( 2016), stepwise regression is still a prevalent practice in published psychological research, even though methodologists have long cautioned that stepwise regression approaches capitalize on sampling error, have poor replicability, and do not correctly identify the best predictor set of a given size (Henderson & Denison, 1989). For example, identifying risk factors of postpartum anxiety and depression (van der Zee-van den Berg et al., 2021), or identifying transdiagnostic factors that predict depression when controlling for demographic factors (Chen et al., 2021). Frequently researchers may be interested in simplifying their model: narrowing the predictor set from a set of candidate predictors, a process called variable selection. does the relationship between cognition and brain network connectivity differ by age, controlling for sex and head motion?). The choice of predictors may be completely pre-specified according to a specific research question (e.g. Researchers must determine which predictors should be included in x. With intercept \(\beta _0\), regression coefficients \(\beta _1., \beta _p\), and residual error variance \(\sigma ^2\). We recommend SSVS as a flexible framework that is well-suited for the field, discuss limitations, and suggest directions for future development. SSVS as investigated here is reasonably computationally efficient and powerful to detect moderate effects in small sample sizes (or small effects in moderate sample sizes), while protecting against false inclusion and without over-penalizing true effects. We demonstrate these advantages and contrast SSVS with lasso type penalization in an application to predict depression symptoms in a large sample and an accompanying simulation study. We investigate the effects of sample size, effect size, and patterns of correlation among predictors on rates of correct and false inclusion and bias in the estimates. In particular we highlight advantages of stochastic search variable selection (SSVS), that make it well suited for variable selection applications in psychology. In this paper, we compare the properties of lasso approaches used for variable selection to Bayesian variable selection approaches. However, several recognized limitations of lasso regularization may limit its suitability for psychological research. ![]() Modern regularization methods such as lasso regression have recently been introduced in the field and are incorporated into popular methodologies, such as network analysis. In the current paper, we review existing tools for solving variable selection problems in psychology. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |