Testing Variance Components in Linear Mixed Modeling Using Permutation

Abstract

Inference of variance components in linear mixed modeling (LMM) provides evidence of heterogeneity between individuals or clusters. When only nonnegative variances are allowed, there is a boundary (i.e., 0) in the variances’ parameter space, and regular inference statistical procedures for such a parameter could be problematic. The goal of this article is to introduce a practically feasible permutation method to make inferences about variance components while considering the boundary issue in LMM. The permutation tests with different settings (i.e., constrained vs. unconstrained estimation, specific vs. generalized test, different ways of calculating p values, and different ways of permutation) were examined with both normal data and non-normal data. In addition, the permutation tests were compared to likelihood ratio (LR) tests with a mixture of chi-squared distributions as the reference distribution. We found that the unconstrained permutation test with the one-sided p-value approach performed better than the other permutation tests and is a useful alternative when the LR tests are not applicable. An R function is provided to facilitate the implementation of the permutation tests, and a real data example is used to illustrate the application. We hope our results will help researchers choose appropriate tests when testing variance components in LMM.     

The Incremental Utility of Emotion Regulation But Not Emotion Reactivity in Nonsuicidal Self‐Injury

This study assessed the incremental utility of emotion reactivity and emotion regulation in relation to nonsuicidal self‐injury (NSSI). Participants included 379 college students aged 18–22 who completed self‐report measures of emotion regulation, emotion reactivity, and NSSI. Emotion regulation was significantly related to NSSI both ignoring and controlling for reactivity, but the reverse was not true. Participants' use of NSSI for affect regulation appeared to moderate this relation. Findings support emotion regulation deficits as a target for intervention over and above heightened emotion reactivity, especially in those who use NSSI to regulate negative affect.     

Suicide Risk Among Wounded U.S. Service Members

The association between suicide and combat injuries sustained during the wars in Iraq and Afghanistan was examined. A retrospective population‐based cohort design was conducted using official military records to identify combat injuries (October 7, 2001, to December 31, 2007). Those who were injured during combat had higher crude suicide rates than those who deployed and were not injured (incidence rate ratio [IRR] = 1.50; confidence interval [CI] = 1.06, 2.12), or never deployed (IRR = 1.46; CI = 1.04, 2.06). After adjusting for demographics, these findings were no longer statistically significant. Although our data did not support an elevated suicide risk among wounded service members, additional research is needed to examine the impact of injury severity.     

The role of pattern recognition in children's exact enumeration of small numbers

Enumeration can be accomplished by subitizing, counting, estimation, and combinations of these processes. We investigated whether the dissociation between subitizing and counting can be observed in 4‐ to 6‐year‐olds and studied whether the maximum number of elements that can be subitized changes with age. To detect a dissociation between subitizing and counting, it is tested whether task manipulations have different effects in the subitizing than in the counting range. Task manipulations concerned duration of presentation of elements (limited, unlimited) and configuration of elements (random, line, dice). In Study 1, forty‐nine 4‐ and 5‐year‐olds were tested with a computerized enumeration task. Study 2 concerned data from 4‐, 5‐, and 6‐year‐olds, collected with Math Garden, a computer‐adaptive application to practice math. Both task manipulations affected performance in the counting, but not the subitizing range, supporting the conclusion that children use two distinct enumeration processes in the two ranges. In all age groups, the maximum number of elements that could be subitized was three. The strong effect of configuration of elements suggests that subitizing might be based on a general ability of pattern recognition.     

It’s a journey: from media effects to dynamic systems

ABSTRACT

This paper aims to demonstrate how an analytical paradigm shift from the General Linear Model (GLM) used in most communication processes and effects research to dynamic systems theory (DST, a nonlinear mathematical theory), fundamentally changes one’s research assumptions and research questions and leads to novel approaches to research design, data collection, and analysis. Concrete examples demonstrating these changes are drawn from the co-viewing literature. In addition, we discuss how data collected and interpreted using the GLM can be re-analyzed and re-interpreted to further inform our understanding of communication behavior when we use the assumptions of dynamic systems theory to derive new predictions.     

Psychosocial–Environmental Risk Factors for Suicide Attempts in Adolescents with Suicidal Ideation: Findings from a Sample of 73,238 Adolescents

We determined risk factors that discriminate between suicide attempt (SA) adolescents and suicidal ideation only (SI only) adolescents using data from the 2010 Korea Youth Risk Behavior Web‐based Survey (12–19 years; N = 73,238). In males, heavy alcohol use, drug use, and high perceived sadness/hopelessness showed significant effects on the presence of SA versus the presence of SI only. In females, along with these variables, low academic achievement, poor perceived health status, high perceived stress, and unhealthy coping strategy were also significantly related to the presence of SA versus SI only. Therefore, clinical interventions targeting adolescents' psychological distress are warranted to prevent suicide.