Department of Medicine Statistics Core
Overview
The Department of Medicine Statistics Core (DOMStat) comprises a team of 35 seasoned statisticians proficient in providing top-tier statistical support to researchers within the Department of Medicine. Our core faculty engages in collaborative efforts with investigators across various phases of research endeavors, including grant preparation, data analysis, and manuscript composition.
Leadership
Services
Consulting Services
- Short-Term Consulting: This service is perfect for investigators who need focused, brief assistance with statistical analysis but do not require long-term collaboration. Eligible individuals include faculty, clinical instructors, fellows, postdocs, students, research staff, and residents.
- Long-Term Consulting: A percentage of skilled personnel’s time can be contracted to work directly with a research team throughout the course of the project. This arrangement is ideal for large-scale studies, multi-phase research, or projects that need consistent statistical oversight.
Database Design & Management
Database Design and Development
Data Entry Supervision: We provide guidance on how to structure data entry forms and oversee data entry processes to ensure consistency and accuracy.
Data Quality Checking: Our team can implement procedures to monitor and verify the accuracy of your data at each stage of collection and entry.
Data Cleaning: We assist with identifying and rectifying any discrepancies, missing values, or outliers in your dataset, ensuring that your data is reliable and ready for analysis.
Interim Reporting: We support the preparation of interim reports, summarizing the status of your data and helping you stay on track with your research objectives.
Data Management Services
Data Entry Supervision: We provide guidance on how to structure data entry forms and oversee data entry processes to ensure consistency and accuracy.
Data Quality Checking: Our team can implement procedures to monitor and verify the accuracy of your data at each stage of collection and entry.
Data Cleaning: We assist with identifying and rectifying any discrepancies, missing values, or outliers in your dataset, ensuring that your data is reliable and ready for analysis.
Interim Reporting: We support the preparation of interim reports, summarizing the status of your data and helping you stay on track with your research objectives.
Grant & Manuscript Preparation
Grant Preparation
- Evaluating Alternative Study Designs: Assistance with selecting the most appropriate study design based on research questions, objectives, and available resources.
- Power and Sample Size Calculations: Guidance on accurate power calculations to ensure studies are properly powered to detect meaningful differences, including determining the correct sample size for reliable results.
- Statistical Analysis Plans: Development of detailed statistical analysis plans outlining the methods and techniques for data analysis.
- Data Collection and Management Strategies: Recommendations on structuring data collection and management plans to maintain data integrity and consistency throughout the research process.
- Randomization and Survey Sampling Approaches: Support designing randomization protocols and selecting appropriate survey sampling methods to minimize bias and enhance study validity
Manuscript Preparation
- Statistical Writing Support:DOMStat statisticians assist in drafting the statistical methods and results sections of manuscripts. They also prepare journal-ready tables and figures that clearly and effectively present study data.
- Manuscript and Abstract Review: The team reviews the statistical content of manuscripts and abstracts to ensure that the statistical methods and results are accurately and clearly described.
- Reviewer Response Assistance: Support is provided in preparing responses to statistical reviewers, helping to address any concerns or clarifications raised during the peer review process.
- Authorship and Collaboration: DOMStat values collaboration and ensures that personnel involved in research projects receive appropriate recognition through authorship on manuscripts. Authorship follows the International Committee of Medical Journal Editors (ICMJE) guidelines.
Statistical Analysis & Programming
Bioinformatics, Biomarkers, and Survival Data Analysis
- High-Throughput -Omics Methods: We assist with the analysis of high-dimensional genomic, transcriptomic, proteomic, and metabolomic data, utilizing advanced techniques to uncover meaningful biological insights.
Statistical Genetics: We provide support for genetic studies, including gene association studies, genome-wide association studies (GWAS), and other genetic analyses, ensuring proper handling of large-scale genetic data.
Biomarker Development and Validation: Our team helps identify and validate biomarkers, using statistical approaches to assess their diagnostic, prognostic, or therapeutic potential.
Survival Analysis: We offer expertise in survival analysis, including Cox proportional hazards models, Kaplan-Meier curves, and time-to-event analysis, to assess and interpret time-dependent outcomes.
Joint Modeling of Longitudinal and Survival Data: We assist in modeling both longitudinal data and survival outcomes simultaneously, such as the relationship between repeated measures of biomarkers and time-to-event outcomes.
Longitudinal, Hierarchical, and Causal Analysis
Longitudinal Analysis: Our statisticians assist in analyzing repeated measures and data collected over time, using techniques such as mixed-effects models and growth curve analysis.
Multilevel/Hierarchical Models: We support the analysis of data with hierarchical or nested structures (e.g., patients within hospitals, students within schools), using multilevel models to account for the dependence between observations.
Quasi-Experimental Designs: We provide statistical expertise in evaluating quasi-experimental research designs, including propensity score matching and instrumental variable analysis, to draw causal inferences from non-randomized studies.
Complex Survey Samples: We support the analysis of complex survey data, using sampling weights, stratification, and clustering to ensure accurate estimation of population parameters.
Advanced Statistical Modeling and Machine Learning
Structural Equation Modeling (SEM): Our team is experienced in using SEM to model complex relationships between observed and latent variables, assessing both direct and indirect effects.
Bayesian Methods: For complex data and models, we apply Bayesian methods to incorporate prior information, estimate uncertainty, and make predictions.
Monte Carlo Simulations: We apply Monte Carlo simulations to model complex systems, estimate uncertainties, and explore various possible scenarios under different assumptions.
Prediction Modeling and Machine Learning: We implement various prediction modeling techniques, including regression models, random forests, and other machine learning algorithms, to forecast future outcomes based on your data.
Data Quality, Reliability, and Specialized Techniques
Psychometric Reliability: We offer support in evaluating the reliability and validity of measurement instruments, using techniques such as factor analysis and Cronbach's alpha.
Imputation and Missing Data Strategies: We provide strategies for handling missing data, including multiple imputation and other advanced methods to minimize bias and maximize the accuracy of your results.
Nonparametric Methods: When assumptions of normality are not met, we implement nonparametric techniques for comparing groups or assessing relationships between variables.
Meta-Analysis: Our team can conduct meta-analyses, synthesizing results from multiple studies to provide an overall estimate of effect and assess study heterogeneity.