Using Lasso for Predictor Selection and to Assuage Overfitting: A Method Long Overlooked in Behavioral Sciences

By Daniel M. McNeish. Multivariate Behavioral Research, 2015

392 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/ba10a616715b79b73362e969f738230d14be6830

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

Citations in the corpus

Extending the Minority Stress Model to Understand Mental Health Problems Experienced by the Autistic Population (2018, methodology): “26 be preferable to ordinary least squares (OLS) regression (Helwig 2017; Jacobucci, Grimm, and McArdle 2016; McNeish 2015).”
Choosing prediction over explanation in psychology: Lessons from machine learning (2017, background): “Although a full treatment of these issues is beyond the scope of this paper (for excellent full-length introductions, see Browne, 2000; Bunea et al., 2011; Domingos, 2012; McNeish, 2015; an authoritative, freely available textbook is Hastie, Tibshirani, & Friedman, 2009), even a cursory familiarity ”
Development of Indirect Measures of Conscientiousness: Combining a Facets Approach and Network Analysis (2015, methodology): “In network analysis performed with the adaptive lasso method, the presence or absence of a relationship is established in a data-dependent manner; therefore, this method does not include the same degrees of freedom while still providing a parsimonious account of the relationships among the variables”
Small Sample Size Solutions (2020, result): “We would be able to fit our linear regression model, but we would be overfitting our model to that specific sample and our results would not generalize well to a different sample from the population (McNeish, 2015).”
A Bayesian Approach for Estimating Multilevel Latent Contextual Models (2016, methodology): “For example, penalized likelihood methods (e.g., Chung et al., 2013; DeCarlo et al., 2011; Galindo-Garre, Vermunt, & Bergsma, 2004;McNeish, 2015;Mislevy, 1986) employweakly informative priors for variance components to avoid boundary estimates (e.g., zero variances; Chung et al., 2013).”