Inferences from logistic regression models in the presence of small samples, rare events, nonlinearity, and multicollinearity with observational data

Bergtold, Jason S.; Yeager, Elizabeth A.; Featherstone, Allen M.

doi:10.1080/02664763.2017.1282441

Cited by 39 publications

(28 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although these analyses do not generate false detection of CNDD, they bias the quantification of both CNDD and HNDD by introducing systematic errors that produce spurious interspecific patterns. For example, logistic regressions, a common tool for survival analysis, are quite sensitive to measuring errors and nonlinearity (Stefansky & Carroll ; Heid et al ; Bergtold et al ). A logistic regression assumes that survival approaches zero for high values of local density.…”

Section: Discussionmentioning

confidence: 99%

Bias in the detection of negative density dependence in plant communities

et al. 2019

View full text Add to dashboard Cite

Regression dilution is a statistical inference bias that causes underestimation of the strength of dependency between two variables when the predictors are error‐prone proxies (EPPs). EPPs are widely used in plant community studies focused on negative density‐dependence (NDD) to quantify competitive interactions. Because of the nature of the bias, conspecific NDD is often overestimated in recruitment analyses, and in some cases, can be erroneously detected when absent. In contrast, for survival analyses, EPPs typically cause NDD to be underestimated, but underestimation is more severe for abundant species and for heterospecific effects, thereby generating spurious negative relationships between the strength of NDD and the abundances of con‐ and heterospecifics. This can explain why many studies observed rare species to suffer more severely from conspecific NDD, and heterospecific effects to be disproportionally smaller than conspecific effects. In general, such species‐dependent bias is often related to traits associated with likely mechanisms of NDD, which creates false patterns and complicates the ecological interpretation of the analyses. Classic examples taken from literature and simulations demonstrate that this bias has been pervasive, which calls into question the emerging paradigm that intraspecific competition has been demonstrated by direct field measurements to be generally stronger than interspecific competition.

show abstract

Section: Discussionmentioning

confidence: 99%

Bias in the detection of negative density dependence in plant communities

et al. 2019

View full text Add to dashboard Cite

show abstract

“…To examine the change in the probability of having a psychiatric disorder over time (i.e., disorder trajectory) and to examine how disorder trajectory related to the various risk factor(s), all MI-GEE analyses included an Age × Risk Factor interaction term as an independent variable, and the interaction effect was illustrated with marginal effects of the independent variables. Marginal effect is robust to sample size and has been shown to be a more appropriate measure of how variables interact in relation to the probability of the outcome variable than the coefficient of the interaction term in the regression (Bergtold, Yeager, & Featherstone, 2011; Berry, DeMeritt, & Esarey, 2010). The slope of the disorder trajectory was reflected by average marginal effects of age (ME A ), the estimated population differences in the probability of having a psychiatric disorder between the two ages.…”

Section: Methodsmentioning

confidence: 99%

Mental health trajectories from adolescence to adulthood: Language disorder and other childhood and adolescent risk factors

2015

View full text Add to dashboard Cite

Longitudinal research on mental health development beyond adolescence among nonclinical populations is lacking. This study reports on psychiatric disorder trajectories from late adolescence to young adulthood in relation to childhood and adolescent risk factors. Participants were recruited for a prospective longitudinal study tracing a community sample of 5-year-old children with communication disorders and a matched control cohort to age 31. Psychiatric disorders were measured at ages 19, 25, and 31. Known predictors of psychopathology and two school-related factors specifically associated with language disorder (LD) were measured by self-reports and semistructured interviews. The LD cohort was uniquely characterized by a significantly decreasing disorder trajectory in early adulthood. Special education was associated with differential disorder trajectories between LD and control cohorts, whereas maltreatment history, specific learning disorder, family structure, and maternal psychological distress were associated with consistent trajectories between cohorts. From late adolescence to young adulthood, childhood LD was characterized by a developmentally limited course of psychiatric disorder; maltreatment was consistently characterized by an elevated risk of psychiatric disorder regardless of LD history, whereas special education was associated with significantly decreasing risk of psychiatric disorder only in the presence of LD.

show abstract

“…All nondichotomous variables have been standardised prior to running the regression analyses (Di Stefano et al 2009: 5-6). Given our relatively small sample size (N ¼ 705), 7 we present marginal effects, instead of odd ratios, to make the interpretation more reliable (Bertgold et al 2018).…”

Section: Data and Variablesmentioning

confidence: 99%

Who votes for new parties? Economic voting, political ideology and populist attitudes

Marcos-Marné

Plaza-Colodro

Freyburg

2019

West European Politics

View full text Add to dashboard Cite

Beginning with the economic crisis in 2008, a number of European societies witnessed the emergence of 'new parties'. Most authors explain their electoral appeal by focusing on how the perceived state of the economy influences individuals' voting decisions. This article determines the extent to which political attitudes can also explain voting for new political parties born in the heat of the economic crisis. Specifically, it explores the link between populist attitudes, in contrast to pluralist and elitist attitudes, and voting for two new political parties in Spain (Ciudadanos; Podemos), which are noticeably different in their ideological positions, programmatic proposals and populist discourses. The results show that stronger populist attitudes increase the likelihood of voting for new parties as dissimilar as Podemos and Ciudadanos. Overall, the findings suggest that voting for new parties cannot be understood as a mere economic response. Rather, political factors, and especially populist attitudes, matter too.

show abstract

Inferences from logistic regression models in the presence of small samples, rare events, nonlinearity, and multicollinearity with observational data

Cited by 39 publications

References 19 publications

Bias in the detection of negative density dependence in plant communities

Bias in the detection of negative density dependence in plant communities

Mental health trajectories from adolescence to adulthood: Language disorder and other childhood and adolescent risk factors

Who votes for new parties? Economic voting, political ideology and populist attitudes

Contact Info

Product

Resources

About