A new goodness-of-fit measure for probit models: Surrogate R2.

Probit models are used extensively for inferential purposes in the social sciences as discrete data are prevalent in a vast body of social studies. Among many accompanying model inference problems, a critical question remains unsettled: how to develop a goodness-of-fit measure that resembles the ordinary least square (OLS) R2 used for linear models. Such a measure has long been sought to achieve 'comparability' of different empirical models across multiple samples addressing similar social questions. To this end, we propose a novel R2 measure for probit models using the notion of surrogacy - simulating a continuous variable S as a surrogate of the original discrete response (Liu & Zhang, Journal of the American Statistical Association, 113, 845 and 2018). The proposed R2 is the proportion of the variance of the surrogate response explained by explanatory variables through a linear model, and we call it a surrogate R2 . This paper shows both theoretically and numerically that the surrogate R2 approximates the OLS R2 based on the latent continuous variable, preserves the interpretation of explained variation, and maintains monotonicity between nested models. As no other pseudo R2 , McKelvey and Zavoina's and McFadden's included, can meet all the three criteria simultaneously, our measure fills this crucial void in probit model inference.

MrIML: Multi-response interpretable machine learning to model genomic landscapes.

We introduce a new R package "MrIML" ("Mister iml"; Multi-response Interpretable Machine Learning). MrIML provides a powerful and interpretable framework that enables users to harness recent advances in machine learning to quantify multilocus genomic relationships, to identify loci of interest for future landscape genetics studies, and to gain new insights into adaptation across environmental gradients. Relationships between genetic variation and environment are often nonlinear and interactive; these characteristics have been challenging to address using traditional landscape genetic approaches. Our package helps capture this complexity and offers functions that fit and interpret a wide range of highly flexible models that are routinely used for single-locus landscape genetics studies but are rarely extended to estimate response functions for multiple loci. To demonstrate the package's broad functionality, we test its ability to recover landscape relationships from simulated genomic data. We also apply the package to two empirical case studies. In the first, we model genetic variation of North American balsam poplar (Populus balsamifera, Salicaceae) populations across environmental gradients. In the second case study, we recover the landscape and host drivers of feline immunodeficiency virus genetic variation in bobcats (Lynx rufus). The ability to model thousands of loci collectively and compare models from linear regression to extreme gradient boosting, within the same analytical framework, has the potential to be transformative. The MrIML framework is also extendable and not limited to modelling genetic variation; for example, it can quantify the environmental drivers of microbiomes and coinfection dynamics.

Risk Factors for Hearing Decrement Among U.S. Air Force Aviation-Related Personnel.

INTRODUCTION: The purpose of this study was to analyze historical hearing sensitivity data to determine factors associated with an occupationally significant change in hearing sensitivity in U.S. Air Force aviation-related personnel. METHODS: This study was a longitudinal, retrospective cohort analysis of audiogram records for Air Force aviation-related personnel on active duty during calendar year 2013 without a diagnosis of non-noise-related hearing loss. The outcomes of interest were raw change in hearing sensitivity from initial baseline to 2013 audiogram and initial occurrence of a significant threshold shift (STS) and non-H1 audiogram profile. Potential predictor variables included age and elapsed time in cohort for each audiogram, gender, and Air Force Specialty Code. Random forest analyses conducted on a learning sample were used to identify relevant predictor variables. Mixed effects models were fitted to a separate validation sample to make statistical inferences. RESULTS: The final dataset included 167,253 nonbaseline audiograms on 10,567 participants. Only the interaction between time since baseline audiogram and age was significantly associated with raw change in hearing sensitivity by STS metric. None of the potential predictors were associated with the likelihood for an STS. Time since baseline audiogram, age, and their interaction were significantly associated with the likelihood for a non-HI hearing profile. DISCUSSION: In this study population, age and elapsed time since baseline audiogram were modestly associated with decreased hearing sensitivity and increased likelihood for a non-H1 hearing profile. Aircraft type, as determined from Air Force Specialty Code, was not associated with changes in hearing sensitivity by STS metric.Greenwell BM, Tvaryanas AP, Maupin GM. Risk factors for hearing decrement among U.S. Air Force aviation-related personnel. Aerosp Med Hum Perform. 2018; 89(2):80-86.

The Commander's Wellness Program: Assessing the Association Between Health Measures and Physical Fitness Assessment Scores, Fitness Assessment Exemptions, and Duration of Limited Duty.

INTRODUCTION: Air Force Medical Service health promotions staff have identified a set of evidenced-based interventions targeting tobacco use, sleep habits, obesity/healthy weight, and physical activity that could be integrated, packaged, and deployed as a Commander's Wellness Program. The premise of the program is that improvements in the aforementioned aspects of the health of unit members will directly benefit commanders in terms of members' fitness assessment scores and the duration of periods of limited duty. The purpose of this study is to validate the Commander's Wellness Program assumption that body mass index (BMI), physical activity habits, tobacco use, sleep, and nutritional habits are associated with physical fitness assessment scores, fitness assessment exemptions, and aggregate days of limited duty in the population of active duty U.S. Air Force personnel. METHODS: This study used a cross-sectional analysis of active duty U.S. Air Force personnel with an Air Force Web-based Health Assessment and fitness assessment data during fiscal year 2013. Predictor variables included age, BMI, gender, physical activity level (moderate physical activity, vigorous activity, and muscle activity), tobacco use, sleep, and dietary habits (consumption of a variety of foods, daily servings of fruits and vegetables, consumption of high-fiber foods, and consumption of high-fat foods). Nonparametric methods were used for the exploratory analysis and parametric methods were used for model building and statistical inference. RESULTS: The study population comprised 221,239 participants. Increasing BMI and tobacco use were negatively associated with the outcome of composite fitness score. Increasing BMI and tobacco use and decreasing sleep were associated with an increased likelihood for the outcome of fitness assessment exemption status. Increasing BMI and tobacco use and decreasing composite fitness score and sleep were associated with an increased likelihood for the outcome of limited duty status, whereas increasing BMI and decreasing sleep were associated with the outcome of increased aggregate days of limited duty. The observed associations were in the expected direction and the effect sizes were modest. Physical activity habits and nutritional habits were not observed to be associated with any of the outcome measures. CONCLUSIONS: The Commander's Wellness Program should be scoped to those interventions targeting BMI, composite fitness score, sleep, and tobacco use. Although neither self-reported physical activity nor nutritional habits were associated with the outcomes, it is still worthwhile to include related interventions in the Commander's Wellness Program because of the finding in other studies of a consistent association between the overall number of health risks and productivity outcomes.