Model selection with misspecified spatial covariance structure

Spatial data analysis has become more and more important in the studies of ecology and economics during the last decade. One focus of spatial data analysis is how to select predictors, variance functions and correlation functions. However, in general, the true covariance function is unknown and the working covariance structure is often misspecified. In this paper, our target is to find a good strategy to identify the best model from the candidate set using model selection criteria. This paper is to evaluate the ability of some information criteria (corrected Akaike information criterion, Bayesian information criterion (BIC) and residual information criterion (RIC)) for choosing the optimal model when the working correlation function, the working variance function and the working mean function are correct or misspecified. Simulations are carried out for small to moderate sample sizes. Four candidate covariance functions (exponential, Gaussian, Matern and rational quadratic) are used in simulation studies. With the summary in simulation results, we find that the misspecified working correlation structure can still capture some spatial correlation information in model fitting. When the sample size is large enough, BIC and RIC perform well even if the the working covariance is misspecified. Moreover, the performance of these information criteria is related to the average level of model fitting which can be indicated by the average adjusted R square ( [GRAPHICS] ), and overall RIC performs well.

Intra-cluster correlation structure in longitudinal data analysis: Selection criteria and misspecification tests

Selection criteria and misspecification tests for the intra-cluster correlation structure (ICS) in longitudinal data analysis are considered. In particular, the asymptotical distribution of the correlation information criterion (CIC) is derived and a new method for selecting a working ICS is proposed by standardizing the selection criterion as the p-value. The CIC test is found to be powerful in detecting misspecification of the working ICS structures, while with respect to the working ICS selection, the standardized CIC test is also shown to have satisfactory performance. Some simulation studies and applications to two real longitudinal datasets are made to illustrate how these criteria and tests might be useful.

Working covariance model selection for generalized estimating equations

We investigate methods for data-based selection of working covariance models in the analysis of correlated data with generalized estimating equations. We study two selection criteria: Gaussian pseudolikelihood and a geodesic distance based on discrepancy between model-sensitive and model-robust regression parameter covariance estimators. The Gaussian pseudolikelihood is found in simulation to be reasonably sensitive for several response distributions and noncanonical mean-variance relations for longitudinal data. Application is also made to a clinical dataset. Assessment of adequacy of both correlation and variance models for longitudinal data should be routine in applications, and we describe open-source software supporting this practice.

Modeling strategies in longitudinal data analysis: Covariate, variance function and correlation structure selection

A modeling paradigm is proposed for covariate, variance and working correlation structure selection for longitudinal data analysis. Appropriate selection of covariates is pertinent to correct variance modeling and selecting the appropriate covariates and variance function is vital to correlation structure selection. This leads to a stepwise model selection procedure that deploys a combination of different model selection criteria. Although these criteria find a common theoretical root based on approximating the Kullback-Leibler distance, they are designed to address different aspects of model selection and have different merits and limitations. For example, the extended quasi-likelihood information criterion (EQIC) with a covariance penalty performs well for covariate selection even when the working variance function is misspecified, but EQIC contains little information on correlation structures. The proposed model selection strategies are outlined and a Monte Carlo assessment of their finite sample properties is reported. Two longitudinal studies are used for illustration.

Estimation for the general sample selection models

Consider a general regression model with an arbitrary and unknown link function and a stochastic selection variable that determines whether the outcome variable is observable or missing. The paper proposes U-statistics that are based on kernel functions as estimators for the directions of the parameter vectors in the link function and the selection equation, and shows that these estimators are consistent and asymptotically normal.

Criteria for working-correlation-structure selection in GEE: Assessment via simulation

Efficiency of analysis using generalized estimation equations is enhanced when intracluster correlation structure is accurately modeled. We compare two existing criteria (a quasi-likelihood information criterion, and the Rotnitzky-Jewell criterion) to identify the true correlation structure via simulations with Gaussian or binomial response, covariates varying at cluster or observation level, and exchangeable or AR(l) intracluster correlation structure. Rotnitzky and Jewell's approach performs better when the true intracluster correlation structure is exchangeable, while the quasi-likelihood criteria performs better for an AR(l) structure.

Feeding mode of Australian infants in the first 12 months of life: An assessment against national breastfeeding indicators

Background: In 2011, Australia published a set of 6 population-level indicators assessing breastfeeding, formula use, and the introduction of soft/semisolid/solid foods. Objectives: This study aimed to report the feeding practices of Australian infants against these indicators and determine the predictors of early breastfeeding cessation and introduction of solids. Methods: Mother–infant dyads (N = 1470) were recruited postnatally in 2 Australian capital cities and regional areas of 1 state between February 2008 and March 2009. Demographic and feeding intention data were collected by self-completed questionnaire at infant birth, with feeding practices (current feeding mode, age of breastfeeding cessation, age of formula and/or solids introduction) reported when the infant was between 4 and 7 months of age, and around 13 months of age. Multiple logistic regression was used to determine the predictors of breastfeeding cessation and solids introduction. Results: Although initiation of breastfeeding was almost universal (93.3%), less than half of the infants were breastfed to 6 months (41.7%) and 33.3% were receiving solids by 4 months. Women who were socially disadvantaged, younger, less educated, unpartnered, primiparous, and/or overweight were most likely to have ceased breastfeeding before 6 months of age, and younger and/or less educated women were most likely to have introduced solid food by 4 months of age. Not producing adequate milk was the most common reason provided for cessation of breastfeeding. Conclusion: The feeding behaviors of Australian infants in the first 12 months fall well short of recommendations. Women need anticipatory guidance as to the indicators of breastfeeding success and the tendency of women to doubt the adequacy of their breast milk supply warrants further investigation.

Eliciting and encoding expert knowledge on variable selection into classical or Bayesian species distribution models

The quality of species distribution models (SDMs) relies to a large degree on the quality of the input data, from bioclimatic indices to environmental and habitat descriptors (Austin, 2002). Recent reviews of SDM techniques, have sought to optimize predictive performance e.g. Elith et al., 2006. In general SDMs employ one of three approaches to variable selection. The simplest approach relies on the expert to select the variables, as in environmental niche models Nix, 1986 or a generalized linear model without variable selection (Miller and Franklin, 2002). A second approach explicitly incorporates variable selection into model fitting, which allows examination of particular combinations of variables. Examples include generalized linear or additive models with variable selection (Hastie et al. 2002); or classification trees with complexity or model based pruning (Breiman et al., 1984, Zeileis, 2008). A third approach uses model averaging, to summarize the overall contribution of a variable, without considering particular combinations. Examples include neural networks, boosted or bagged regression trees and Maximum Entropy as compared in Elith et al. 2006. Typically, users of SDMs will either consider a small number of variable sets, via the first approach, or else supply all of the candidate variables (often numbering more than a hundred) to the second or third approaches. Bayesian SDMs exist, with several methods for eliciting and encoding priors on model parameters (see review in Low Choy et al. 2010). However few methods have been published for informative variable selection; one example is Bayesian trees (O’Leary 2008). Here we report an elicitation protocol that helps makes explicit a priori expert judgements on the quality of candidate variables. This protocol can be flexibly applied to any of the three approaches to variable selection, described above, Bayesian or otherwise. We demonstrate how this information can be obtained then used to guide variable selection in classical or machine learning SDMs, or to define priors within Bayesian SDMs.

Marker selection by Akaike information criterion and Bayesian information criterion

We carried out a discriminant analysis with identity by descent (IBD) at each marker as inputs, and the sib pair type (affected-affected versus affected-unaffected) as the output. Using simple logistic regression for this discriminant analysis, we illustrate the importance of comparing models with different number of parameters. Such model comparisons are best carried out using either the Akaike information criterion (AIC) or the Bayesian information criterion (BIC). When AIC (or BIC) stepwise variable selection was applied to the German Asthma data set, a group of markers were selected which provide the best fit to the data (assuming an additive effect). Interestingly, these 25-26 markers were not identical to those with the highest (in magnitude) single-locus lod scores.

The moderating effects of job control and selection, optimization, and compensation strategies on the age-work ability relationship

Work ability describes employees' capability to carry out their work with respect to physical and psychological job demands. This study investigated direct and interactive effects of age, job control, and the use of successful aging strategies called selection, optimization, and compensation (SOC) in predicting work ability. We assessed SOC strategies and job control by using employee self-reports, and we measured employees' work ability using supervisor ratings. Data collected from 173 health-care employees showed that job control was positively associated with work ability. Additionally, we found a three-way interaction effect of age, job control, and use of SOC strategies on work ability. Specifically, the negative relationship between age and work ability was weakest for employees with high job control and high use of SOC strategies. These results suggest that the use of successful aging strategies and enhanced control at work are conducive to maintaining the work ability of aging employees. We discuss theoretical and practical implications regarding the beneficial role of the use of SOC strategies utilized by older employees and enhanced contextual resources at work for aging employees.

The buffering effect of selection, optimization, and compensation strategy use on the relationship between problem solving demands and occupational well-being: A daily diary study

This study investigated within-person relationships between daily problem solving demands, selection, optimization, and compensation (SOC) strategy use, job satisfaction, and fatigue at work. Based on conservation of resources theory, it was hypothesized that high SOC strategy use boosts the positive relationship between problem solving demands and job satisfaction, and buffers the positive relationship between problem solving demands and fatigue. Using a daily diary study design, data were collected from 64 administrative employees who completed a general questionnaire and two daily online questionnaires over four work days. Multilevel analyses showed that problem solving demands were positively related to fatigue, but unrelated to job satisfaction. SOC strategy use was positively related to job satisfaction, but unrelated to fatigue. A buffering effect of high SOC strategy use on the demands-fatigue relationship was found, but no booster effect on the demands-satisfaction relationship. The results suggest that high SOC strategy use is a resource that protects employees from the negative effects of high problem solving demands.

Maintaining a focus on opportunities at work: The interplay between age, job complexity, and the use of selection, optimization, and compensation strategies

The concept of focus on opportunities describes how many new goals, options, and possibilities employees believe to have in their personal future at work. This study investigated the specific and shared effects of age, job complexity, and the use of successful aging strategies called selection, optimization, and compensation (SOC) in predicting focus on opportunities. Results of data collected from 133 employees of one company (mean age = 38 years, SD = 13, range 16–65 years) showed that age was negatively, and job complexity and use of SOC strategies were positively related to focus on opportunities. In addition, older employees in high-complexity jobs and older employees in low-complexity jobs with high use of SOC strategies were better able to maintain a focus on opportunities than older employees in low-complexity jobs with low use of SOC strategies.

Influences on novice drivers' speed selection - a systematic review: Application and limitations of theory and methodology

Background Excessive speed is a primary contributing factor to young novice road trauma, including intentional and unintentional speeds above posted limits or too fast for conditions. The objective of this research was to conduct a systematic review of recent investigations into novice drivers’ speed selection, with particular attention to applications and limitations of theory and methodology. Method Systematic searches of peer-reviewed and grey literature were conducted during September 2014. Abstract reviews identified 71 references potentially meeting selection criteria of investigations since the year 2000 into factors that influence (directly or indirectly) actual speed (i.e., behaviour or performance) of young (age <25 years) and/or novice (recently-licensed) drivers. Results Full paper reviews resulted in 30 final references: 15 focused on intentional speeding and 15 on broader speed selection investigations. Both sets identified a range of individual (e.g., beliefs, personality) and social (e.g., peer, adult) influences, were predominantly theory-driven and applied cross-sectional designs. Intentional speed investigations largely utilised self-reports while other investigations more often included actual driving (simulated or ‘real world’). The latter also identified cognitive workload and external environment influences, as well as targeted interventions. Discussion and implications Applications of theory have shifted the novice speed-related literature beyond a simplistic focus on intentional speeding as human error. The potential to develop a ‘grand theory’ of intentional speeding emerged and to fill gaps to understand broader speed selection influences. This includes need for future investigations of vehicle-related and physical environment-related influences and methodologies that move beyond cross-sectional designs and rely less on self-reports.

Selection, optimization, and compensation strategies: Interactive effects on daily work engagement

The theory of selective optimization with compensation (SOC) proposes that the “orchestrated” use of three distinct action regulation strategies (selection, optimization, and compensation) leads to positive employee outcomes. Previous research examined overall scores and additive models (i.e., main effects) of SOC strategies instead of interaction models in which SOC strategies mutually enhance each other's effects. Thus, a central assumption of SOC theory remains untested. In addition, most research on SOC strategies has been cross-sectional, assuming that employees' use of SOC strategies is stable over time. We conducted a quantitative diary study across nine work days (N = 77; 514 daily entries) to investigate interactive effects of daily SOC strategies on daily work engagement. Results showed that optimization and compensation, but not selection, had positive main effects on work engagement. Moreover, a significant three-way interaction effect indicated that the relationship between selection and work engagement was positive only when both optimization and compensation were high, whereas the relationship was negative when optimization was low and compensation was high. We discuss implications for future research and practice regarding the use of SOC strategies at work.

Born leaders: Political selection and the relative age effect in the US Congress

We present substantial evidence for the existence of a bias in the distribution of births of leading US politicians in favour of those who were the eldest in their cohort at school. This result adds to the research on the long-term effects of relative age among peers at school. We discuss parametric and non-parametric tests to identify this effect, and we show that it is not driven by measurement error, redshirting or a sorting effect of highly educated parents. The magnitude of the effect that we estimate is larger than what other studies on ‘relative age effects’ have found for broader populations but is in general consistent with research that looks at professional sportsmen. We also find that relative age does not seem to correlate with the quality of elected politicians.