Michael Harhay

EPID5360 - Data Mgt&Visualization 1

The objective of this two-course series is to enhance MSCE students' comfort and acumen in all aspects of clinical epidemiological data management and presentation, particularly graphical representation of results. The course progresses from best practices in data collection and database use to advanced data management, summarization of results, and data visualization, all of which are grounded in the prioritization of producing efficient and reproducible research processes. The course will cover and develop skills in: basic data collection, harmonization, and integration with Stata software; best practices for data variable derivation and creation; assessing and dealing with missing data; merging and appending datasets; management of dates and times; assessing free text data; dealing with specific data types such as ICD-9 and 10 codes, cost data, management of longitudinal and time-to-event data; production of descriptive and regression tables (for all regression types); descriptive and regression model visualization; and the use of Stata Markdown files such that research reports can be created directly from Stata. By the end of the two-course series, students will become fluent in the Stata statistical language and be uniquely positioned to advance their independent clinical epidemiological careers through best research and data presentation practices.

EPID5370 - Data Mgt&Visualization 2

The objective of this two-course series is to enhance MSCE students' comfort and acumen in all aspects of clinical epidemiological data management and presentation, particularly graphical representation of results. The course progresses from best practices in data collection and database use to advanced data management, summarization of results, and data visualization, all of which are grounded in the prioritization of producing efficient and reproducible research processes. The course will cover and develop skills in: basic data collection, harmonization, and integration with Stata software; best practices for data variable derivation and creation; assessing and dealing with missing data; merging and appending datasets; management of dates and times; assessing free text data; dealing with specific data types such as ICD-9 and 10 codes, cost data, management of longitudinal and time-to-event data; production of descriptive and regression tables (for all regression types); descriptive and regression model visualization; and the use of Stata Markdown files such that research reports can be created directly from Stata. By the end of the two-course series, students will become fluent in the Stata statistical language and be uniquely positioned to advance their independent clinical epidemiological careers through best research and data presentation practices.

EPID6100 - Tutorial in Epi Research

This is a tutorial given by each student’s MSCE mentor. The mentor and student meet regularly, usually weekly. Topics include discussion and review of epidemiologic concepts and principles, guided readings in the epidemiology of a specific health area, and the development of the research protocol. Credit for this course is awarded upon completion of a research project protocol, the one to be used to fulfill the MSCE thesis requirement, which must be approved by the student’s mentor. Evaluation is based on the grade received for the protocol.

EPID6990 - Lab Rotation

Student lab rotation.

EPID6990A - Lab Rotation

EPID7020 - Adv Tpcs Epid Research

The overarching goal of this course is to expose doctoral students in epidemiology to advanced epidemiologic and statistical research methods and theories that are limitedly or not otherwise covered in courses available in the curriculum. Topics that will be covered include reporting guidelines and best practices for reporting statistical methods and results, handling missing data, purposeful selection and application of propensity scores, selected topics in longitudinal and clustered data analysis, contemporary topics in statistical inference and use of p-values and other Frequentist statistical methods, Bayesian theory and inference, and topics selected in collaboration with students and the Graduate Group in Epidemiology and Biostatistics (GGEB) each term. This course is intended for doctoral students in the PhD program in Epidemiology. However, students from other graduate groups are welcome, as long as they meet the pre-requisites; such students are welcome during any year of study. Three learning objectives have been developed for this course; (i) provide students with an understanding of modern and cutting-edge quantitative methods, advanced topics, and best practices in epidemiologic, statistical, and biomedical research; (ii) develop students competence and confidence in statistical programing to support accurate and reproducible epidemiologic and biostatistical analyses; (iii) improve the ability of students to make informed decisions regarding the selection of analytic methods in their individual and collaborative research projects. This course emphasizes the following core competencies: knowledge within program area (epidemiologic and biostatistical methods); research skills (study planning, critically appraising published research); quantitative and computational methodologies (data manipulation, data analysis, statistical coding and debugging, Bayesian inference, data visualization, purposeful statistical inference, and model selection). Through technical lectures, reading of carefully selected peer-reviewed tutorials, critical appraisal of published research studies, and in-class statistical coding laboratory sessions, this course will provide instruction on rigorous and informed statistical model selection, estimation, and interpretation.

EPID8990 - Pre-Dissertation Lab Rot

EPID9900 - Master's Thesis

These are a series of tutorial sessions conducted by the student’s mentor intended to support the student’s efforts in developing a research protocol, designing a research project, and completing the study.

MTR6280 - Quantitative Methods for LHSS

This course explores core principles, theories, and methods from epidemiology, causal inference, biostatistics, and data science, with an emphasis on their application to inform, address, and evaluate health systems-focused research questions and interventions. The ideal learner will have a general familiarity with data analysis, electronic medical records, and experience or planned health systems projects, and is likely a doctoral student, post-doctoral researcher, or faculty member; however, interested students may contact the professor to discuss. The course will cover a wide range of topics to enhance students' familiarity, literacy, and critical appraisal skills in health-system-based randomized trials (including cluster and pragmatic trials), quasi-experimental and observational study designs and methods (such as pre/post studies, differences-in-differences, and time series analysis), as well as general considerations related to multivariable regression modeling, measurement error, missing data, predictive modeling, data integration, and related and emerging topics in learning health system science. Additional topics will vary yearly based on the availability of guest lectures and student composition and needs and may include, for example, lectures on advanced methods (such as Bayesian statistics for clinical research and machine learning), scientific and grant writing, informed consent ,and research ethics. Classes will be centered around instructor-led lectures, journal clubs, student-led presentations, case studies, and expert panels.