This free monthly webinar features Dr. Paul Kolm, Associate Director in the Center for Biostatistics, Informatics, & Data Science (CBIDS) at the MedStar Health Research Institute.
Topics for the Statistical Seminars 2025-2026:
September 19, 2025 - Classical vs. Bayesian Statistics: In this session, we will compare and contrast classical (frequentist) and Bayesian perspectives of statistical analyses.
October 17, 2025 - Exploratory and Confirmatory Factor Analysis: This seminar will look at factor analysis - using a set of observed variables to define an unobserved variable termed a latent variable. We will consider computation methods, evaluation metrics, and presentation of results.
November 21, 2025 - Resampling Statistics: Classical and Bayesian statistics assume there is an underlying theoretical distribution (e.g., normal, chi-square) that can be used to model the data of interest and thereby obtain probabilities of obtaining the value of a statistical test (e.g., t, F, chi-square). In this session, we will look at distribution-free statistics and probabilities. We will focus on statistics based on resampling methods – the bootstrap and the jackknife.
December 19, 2025: No seminar
January 16, 2026 - Analysis of Count Data: In this session, we will discuss regression methods for analyzing outcome data that are counts of events over a given time period. These include Poisson regression, negative binomial regression, zero-inflated Poisson regression, and hurdle models.
February 20, 2026 - Bivariate Time-Series Analysis: This analysis looks at the relationship between two time series and assess whether one “leads” the other and can be said to “cause” the other. Only one aspect of causality is assessed – that “cause” must come before (lead) the effect. We will also discuss the related notion of “Granger causality”.
March 20, 2026 - Missing Data: Missing data presents a potential threat to study validity. In this session, we will consider assumptions about missingness: missing completely at random (MCAR), missing at random (MAR), and not missing at random (NMAR), and how these determine methods for multiple imputation of missing values. Discussion includes older methods of imputation that are less used but may be appropriate in some situations.
April 17, 2026 - Economic Analysis Alongside Clinical Trials: Randomized clinical trials often include an economic component of cost-effectiveness or cost-savings. We will look at the statistical analyses used to assess cost-effectiveness that include accounting for the variability in health resource costs. This includes the method of probability (or Bayesian) sensitivity analysis.
May 15, 2026 - Longitudinal Data Analysis: In this session, we will consider methods for analyzing data obtained from the same cohort of patients over a given time period, i.e., repeated measures. We will look at repeated measures of continuous outcomes and categorical outcomes.
Register now for the MHRI-GHUCCTS Monthly Statistical Seminar Series 2025-2026: https://georgetown.zoom.us/webinar/register/WN_YOYW52dJRB2fogBq-3el9w