Request Expert

This expert is a Harvard PhD graduate and currently Associate Professor of Biostatistics at a U.S. university, where her position is tenured. Her biostatistical and statistical research and teaching includes applications Immunology and Infectious Diseases; Pathology; Epidemiology; Anesthesiology, Perioperative, and Pain Medication; Neurology; Cardiovascular disease; Psychiatry; and Psychology.

Expert's career focus has been the design, analysis, and methodological issues across a range of clinical and public health areas including cancer epidemiology, cardiovascular disease, neurological disease including stroke and traumatic brain injury, dietary epidemiology, and mental health and substance.

Expert has a solid understanding and demonstrated expertise in general statistical methodology, and especially methodological issues in above-mentioned areas, including longitudinal data analysis, survey design and analysis, psychometrics, analysis of large observational data bases using state-of-the art causal inference techniques, advanced methods for survival analysis, non-linear models, and non-parametrics, latent variable analysis including structural equation modeling and factor analysis, meta-analysis, systematic reviews, publication bias, validation of self-report measures (e.g., quality of life, dietary intake, substance use reports), scale and medical device validation using receiver operator characteristics. She also has a wide range of academic interests and expertise encompassing both theory and practice in modern statistics. More important, she is driven to constantly learn new tasks and add more training to her skill set.

Expert has published extensively on latent variable models for misclassified outcomes, joint models for longitudinal count data, Bayesian statistical methods and software, methods for correcting information bias, e.g., misclassified outcomes/co-variates, variable selection, propensity scoring, methods for validating new devices/diagnostics compared to previously established tests, and novel methods for meta-analysis and publication bias. For instance, she developed and published on frequentist and Bayesian methods for latent variable analysis and variable selection to predict cancer subtypes from genetic data, in the presence of outcome information bias (measurement error). She has also published on hidden Markov models, propensity matching methods for confounding, and multivariate methods for outcome reporting bias in a variety of areas, including self-report variables. Her secondary academic interests target statistical methods for bias correction inherent in epidemiological studies -- for example, those collected from clinical convenience samples, surveys, or electronic health records

Overall, Expert has extensive expertise in simulation, model validation, and programming in all statistical languages, using both frequentist and Bayesian paradigms. She is published in the above mentioned areas with over 60 peer-reviewed publications since 2008. She has also taught at a Ph.D level for 12 years (survival analysis, linear regression, linear models, multivariate statistics, and teaching experiences), and is forever eager to take on new challenges.

She is an expert in SAS, R, STATA, SPSS, WinBUGS, JAGS, openBUGS computing environments, and am proficient in SUDAAN, MATLAB, C++ languages and computing environments. She is also conversational in the Spanish language.

Much of the Expert's work involves design and rigorous statistical analysis of randomized controlled trials, as well as the analysis of large epidemiological survey databases such as CDC’s NHANES data base, NIH’s National Longitudinal Survey of Youth, and similarly designed surveys that incorporate complex sampling weights. Several of the large randomized controlled trials she has designed and analyzed also employ complex clustered designs similar to these stratified clustered surveys.

Expert's epidemiologic work has specifically focused on cancer, substance use, and cardiovascular disease, e.g., determining risk factors for these diseases from these large, observational data bases. For such studies, in addition to standard classical statistical analyses, Expert has developed and applied data mining and state-of-the-art frequentist and Bayesian clustering and variable selection methods. She has written papers on misclassification of outcomes, and utilized joint modeling approach for missing data. She has used similar data mining methods to analyze large medical records data bases to determine pharmacological agents that reduce cancer incidence. She has developed and compared missing data algorithms for large epidemiologic genetic studies, and has utilized causal inference techniques such as propensity matching to draw inference from observational studies with extensive confounding. Her additional areas of interest in epidmiology include fetal and maternal health, environmental risk factors for disease, and alcohol and substance use research. She is published in the above-mentioned areas including in high-impact publications such as JAMA, Circulation, Archives of General Psychiatry, Critical Care Medicine, Drug and Alcohol Dependence, NeuroImage, PLOS One, Biometrics, Biostatistics, Journal of the Royal Statistical Society, Statistical Methods in Medical Research, Medical Decision Making, and Statistics in Medicine. She has served as a reviewer for 25 journals and have served as a scientific reviewer for National Institute of Health (NIH) and the Department of Defense. She has been the recipient of multiple highly competitive research grants from the NIH to research the above-mentioned topics.