Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Statistical modeling methods: challenges and strategies

Statistical modeling methods: challenges and strategies Statistical modeling methods are widely used in clinical science, epidemiology, and health services research to analyze data that has been collected in clinical trials as well as observational studies of existing data sources, such as claims files and electronic health records. Diagnostic and prognostic inferences from statistical models are critical to researchers advancing science, clinical practitioners making patient care decisions, and administrators and policy makers impacting the health care system to improve quality and reduce costs. The veracity of such inferences relies not only on the quality and completeness of the collected data, but also statistical model validity. A key component of establishing model validity is determining when a model is not correctly specified and therefore incapable of adequately representing the Data Generating Process (DGP). In this article, model validity is first described and methods designed for assessing model fit, specification, and selection are reviewed. Second, data transformations that improve the model’s ability to represent the DGP are addressed. Third, model search and validation methods are discussed. Finally, methods for evaluating predictive and classification performance are presented. Together, these methods provide a practical framework with recommendations to guide the development and evaluation of statistical models that provide valid statistical inferences. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Biostatistics & Epidemiology Taylor & Francis

Statistical modeling methods: challenges and strategies

Download PDF

Statistical modeling methods: challenges and strategies

Abstract

Statistical modeling methods are widely used in clinical science, epidemiology, and health services research to analyze data that has been collected in clinical trials as well as observational studies of existing data sources, such as claims files and electronic health records. Diagnostic and prognostic inferences from statistical models are critical to researchers advancing science, clinical practitioners making patient care decisions, and administrators and policy makers impacting the...
Loading next page...
 
/lp/taylor-francis/statistical-modeling-methods-challenges-and-strategies-o4RUlWi5Co
Publisher
Taylor & Francis
Copyright
© 2019 Martingale Research Corporation
ISSN
2470-9379
eISSN
2470-9360
DOI
10.1080/24709360.2019.1618653
Publisher site
See Article on Publisher Site

Abstract

Statistical modeling methods are widely used in clinical science, epidemiology, and health services research to analyze data that has been collected in clinical trials as well as observational studies of existing data sources, such as claims files and electronic health records. Diagnostic and prognostic inferences from statistical models are critical to researchers advancing science, clinical practitioners making patient care decisions, and administrators and policy makers impacting the health care system to improve quality and reduce costs. The veracity of such inferences relies not only on the quality and completeness of the collected data, but also statistical model validity. A key component of establishing model validity is determining when a model is not correctly specified and therefore incapable of adequately representing the Data Generating Process (DGP). In this article, model validity is first described and methods designed for assessing model fit, specification, and selection are reviewed. Second, data transformations that improve the model’s ability to represent the DGP are addressed. Third, model search and validation methods are discussed. Finally, methods for evaluating predictive and classification performance are presented. Together, these methods provide a practical framework with recommendations to guide the development and evaluation of statistical models that provide valid statistical inferences.

Journal

Biostatistics & Epidemiology – Taylor & Francis

Published: Jan 1, 2020

Keywords: Goodness-of-fit; Information Matrix Test; model misspecification; model selection; specification analysis

References

  • Applied regression analysis
  • Model transparency and validation: a report of the ISPOR-SMDM modeling good research practices task force–7
  • Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement
  • Improving transparency and replication in Bayesian statistics: the WAMBS-checklist
  • The misuse and abuse of statistics in biomedical research
  • Modeling good research practices–overview: a report of the ISPOR-SMDM modeling good research practices task force–1
  • An introduction to variable and feature selection
  • Consequences of model misspecification for maximum likelihood estimation with missing data
  • Challenges and strategies in analysis of missing data
  • Making causal inferences about treatment effect sizes from observational datasets
  • Causal diagrams for empirical research
  • Directed acyclic graphs, sufficient causes, and the properties of conditioning on a common effect
  • Bayesian model averaging: a tutorial
  • Statistical conclusion validity: some common threats and simple remedies
  • Multimodel inference: understanding AIC and BIC in model selection
  • The importance of external validity
  • External validity: the neglected dimension in evidence ranking
  • External validity: from do-calculus to transportability across populations
  • Assessing the generalizability of prognostic information
  • What do we mean by validating a prognostic model?
  • How to assess the external validity and model validity of therapeutic trials: a conceptual approach to systematic review methodology
  • Internal validation of predictive models: efficiency of some procedures for logistic regression analysis
  • Early blood product and crystalloid volume resuscitation: risk association with multiple organ dysfunction after severe blunt traumatic injury
  • Comparing clinical predictors of deep venous thrombosis vs. pulmonary embolus after severe blunt injury: a new paradigm for post-traumatic venous thromboembolism?
  • A systematic approach to subgroup analyses in a smoking cessation trial
  • Response to letter regarding “a systematic approach to subgroup analyses in a smoking cessation trial”
  • Improving the analysis and modeling of substance use
  • Impact of attention-deficit/hyperactivity disorder (ADHD) treatment on smoking cessation intervention in ADHD smokers: a randomized, double-blind, placebo-controlled trial
  • Generalized information matrix tests for detecting model misspecification
  • New directions in information matrix testing: eigenspectrum tests
  • Maximum likelihood estimation of misspecified models
  • Log-likelihood-based pseudo-R2 in logistic regression: deriving sample-sensitive benchmarks
  • On information and sufficiency
  • Regression and time series model selection in small samples
  • A corrected Akaike information criterion for vector autoregressive model selection
  • Estimating the dimension of a model
  • Bayesian predictive information criterion for the evaluation of hierarchical Bayesian and empirical Bayes models
  • Predictive Bayesian model selection
  • Akaike’s information criterion and recent developments in information complexity
  • Distribution of information statistics and a criterion of model fitting for adequacy of models
  • Asymptotic MAP criteria for model selection
  • Model selection principles in misspecified models
  • Precision, complexity and Bayesian model determination
  • A large-sample model selection criterion based on Kullback’s symmetric divergence
  • A note on overfitting properties of KIC and KICc
  • A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity
  • Consequences and detection of misspecified nonlinear regression models
  • On the consequences of model misspecification in logistic regression
  • Model specification: the views of Fisher and Neyman, and later developments
  • Regression model diagnostics
  • Detection of influential observation in linear regression
  • A survey of outlier detection methodologies
  • The large-sample distribution of the likelihood ratio for testing composite hypotheses
  • Detecting non-linearity in time series
  • The importance of assessing the fit of logistic regression models: a case study
  • A comparison of goodness-of-fit tests for the logistic regression model
  • Goodness-of-fit processes for logistic regression: simulation results
  • Global goodness-of-fit tests in logistic regression with sparse data
  • Pythagorean generalization of testing the equality of two symmetric positive definite matrices
  • Generalized information matrix tests for copulas
  • A goodness-of-fit test for copulas
  • Efficient information based goodness-of-fit tests for vine copula models with fixed margins: a comprehensive review
  • A goodness-of-fit test for regular vine copula models
  • The information matrix test for the linear model
  • Generalized linear models
  • Making correct statistical inferences using a wrong probability model
  • Goodness-of-fit test for a logistic regression model fitted using survey sample data
  • Unweighted sum of squares test for proportions
  • An improved goodness-of-test for logistic regression models based on case-control data by random partition
  • Goodness-of-fit testing for multiple logistic regression analysis when the estimated probabilities are small
  • A goodness-of-fit test for the multiple logistic regression model
  • A goodness-of-fit test for logistic regression models based on case-control data
  • A note on a goodness-of-fit test for the logistic regression model
  • A chi-squared goodness-of-fit test for logistic regression models based on case-control data
  • Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors
  • Simple test for heteroscedasticity and random coefficient variation
  • Testing for serial correlation in least squares regression
  • The Durbin-Watson test for autocorrection in nonlinear models
  • On the formulation of empirical models in dynamic econometrics
  • Standardizing the power of the Hosmer-Lemeshow goodness of fit test in large data sets
  • On a measure of a lack of fit in time series models
  • Discrepancy risk model selection test theory for comparing possibly misspecified or nonnested models
  • Likelihood ratio tests for model selection and non-nested hypotheses
  • Statistical tests for comparing possibly misspecified and nonnested models
  • Model selection tests for nonlinear dynamic models
  • Using bootstrap likelihood ratios in finite mixture models
  • Power analysis for the bootstrap likelihood ratio test for the number of classes in latent class models
  • A bootstrap generalized likelihood ratio test in discriminant analysis
  • On bootstrapping the likelihood ratio test statistic for the number of components in a normal mixture
  • Testing the number of components in a normal mixture
  • Asymptotic theory of certain “goodness-of-fit” criteria based on stochastic processes
  • A multivariate Kolmogorov–Smirnov test of goodness of fit
  • Sulla Determinazione Empirica di una Legge di Distribuzione
  • Table for estimating the goodness of fit of empirical distributions
  • A Kolmogorov-Smirnov test to detect changes in stationarity in big data
  • An analysis of variance test for normality (complete samples)
  • A new measure of rank correlation
  • A data accounting system for clinical investigators
  • An analysis of transformations
  • Assessing the preventive effects of cognitive therapy following relief of depression: a methodological innovation
  • Studying the effects of ACGME duty hours limits on resident satisfaction: results from VA learners’ perceptions survey
  • Do interactions matter?
  • Interaction effects in econometrics
  • Interaction
  • Tests for interaction in epidemiologic studies: a review and a study of power
  • Computational procedures for probing interactions in OLS and logistic regression: SPSS and SAS implementations
  • Relative power of moderated multiple regression and the comparison of subgroup correlation coefficients for detecting moderator effects
  • Model uncertainty
  • Assessment and propagation of model uncertainty
  • Modeling and variable selection in epidemiologic analysis
  • The variable selection problem
  • Marginal asymptotics for the “large p, small n” paradigm: with applications to microarray data
  • Sure independence screening for ultrahigh dimensional feature space
  • Tuning parameter selection in high dimensional penalized likelihood
  • High-dimensional variable selection
  • Bayesian factor regression models in the “large p, small n” paradigm
  • Statistical challenges of high-dimensional data
  • Regression shrinkage and selection via the LASSO
  • Multiple regression analysis
  • All possible regressions with less computation
  • Regression by leaps and bounds
  • Branch-and-bound algorithms for computing the best-subset regression models
  • Branch and bound in statistical data
  • Shotgun stochastic search for “large p” regression
  • Variable selection in linear regression with many predictors
  • Variable selection via Gibbs sampling
  • Markov chain Monte Carlo methods for computing Bayes factors: a comparative review
  • The analysis and selection of variables in linear regression
  • The best sub-set in multiple-regression analysis
  • Efficient calculation of all possible regressions
  • Best subset selection via a modern optimization lens
  • Logistic regression: from art to science
  • Best subsets logistic regression
  • A fast model selection procedure for large families of models
  • A fast procedure for model search in multidimensional contingency tables
  • Efficient screening of nonnormal regression models
  • Computational efficiency in the selection of regression variables
  • Efficient strategies for deriving the subset VAR models
  • An efficient branch-and-bound strategy for subset vector autoregressive model selection
  • Practical Markov chain Monte Carlo
  • Sampling-based approaches to calculating marginal densities
  • Approaches for Bayesian variable selection
  • Model selection and accounting for model uncertainty in graphical models using Occam’s window
  • Risk factors of coronary heart disease: A Bayesian model averaging approach
  • Comparison of Bayesian model averaging and stepwise methods for model selection in logistic regression
  • Variable selection and Bayesian model averaging in case-control studies
  • Model selection in medical research: a simulation study comparing Bayesian model averaging and stepwise regression
  • Bayesian model search and multilevel inference for SNP association studies
  • Bayesian model averaging in proportional hazard models. assessing the risk of a stroke
  • The effects of model selection on confidence-intervals for the size of a closed population
  • Accounting for model uncertainty in estimating global burden of disease
  • Bayesian model averaging: theoretical developments and practical applications
  • Predictive likelihood for Bayesian model selection and averaging
  • Comparison of stepwise selection and Bayesian model averaging for yield gap analysis
  • Imputation and variable selection in linear regression models with missing covariates
  • Accounting for uncertainty in health economic decision models by using model averaging
  • Bayesian model averaging with applications to benchmark dose estimation for arsenic in drinking water
  • Bayesian model averaging and model search strategies
  • Bayesian model averaging for linear regression models
  • Model selection: an integral part of inference
  • Frequentist model averaging estimation: a review
  • Parametric and nonparametric frequentist model selection and model averaging
  • Forensics: assessing model goodness: a machine learning view
  • Cross-validation of regression models
  • Cross-validatory choice and assessment of statistical predictions
  • Mean absolute percentage error for regression models
  • A note on the prediction sum of squares statistic for restricted least squares
  • The relationship between variable selection and data augmentation and a method for prediction
  • A systematic analysis of performance measures for classification tasks
  • An introduction to ROC analysis
  • Basic principles of ROC analyses
  • Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine
  • The meaning and use of the area under a receiver operating characteristic (ROC) curve
  • Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach
  • Logistic regression in rare events data
  • Goodness of prediction fit
  • Predictive fit for natural exponential families
  • Improved predictive model selection
  • Nonparametric bootstrap prediction
  • The Youden index and the optimal cut-point corrected for measurement error
  • Index for rating diagnostic tests
  • Assessing the accuracy of prediction algorithms for classification: an overview
  • A coefficient of agreement for nominal scales
  • Comparison of the predicted and observed secondary structure of T4 phage lysozyme
  • Evaluation: from precision, recall and F-measure to ROC, informedness, markedness & correlation
  • Optimal choice of a cut point for a quantitative diagnostic test performed for research purposes