The UCSF Department of Epidemiology and Biostatistics is launching a series of new short-course modules, as part of a new program titled “Training in Reproducible Research on Aging for Social Science and Epidemiology (TRASE)." There will be 4 modules:
- Data Analysis Skills for Reproducible Social and Behavioral Research on Health & Aging (Spring 2024)
- Reproducibility Core Concepts (date TBA)
- Reproducible Research Skills for Evidence Synthesis, Meta-science and Triangulation (Date TBA)
- Reproducible Research Skills for Data Collection for Social and Behavioral Research on Health & Aging (TBD 2024)
TRASE aims to improve the reproducibility of research on health disparities and aging. The program will offer four short-courses (each can be taken independently) that cover different aspects of reproducible research, from conceptual to technical. The modules are designed for researchers and consumers of scientific research who want to learn how to evaluate, implement, collect, and integrate evidence in a transparent and rigorous way.
The modules are short (~3-6 days) and intensive, combining lectures (recorded or in-person), interactive activities, and hands-on exercises. Participants can choose to attend one or more modules depending on their needs and interests. The TRASE program leverages the expertise and experience of the UCSF faculty and staff who have successfully developed and delivered other training programs in this field.
The Spring 2024 Module registration and waitlist are closed.
Please bookmark this webpage for upcoming short course module offerings and sign up for our TRASE listserv.
Data Analysis Skills for Reproducible Social and Behavioral Research
COURSE INSTRUCTORS
- Fei Jiang, PhD, Associate Professor, Epidemiology & Biostatistics
- Aaron Scheffler, PhD, Assistant Professor, Epidemiology & Biostatistics
Course Description: Concerns around bias and mistakes have been central to the reproducible research movement. Bias includes employing inappropriate statistical methods, model misspecification, and unintentional P-hacking. Mistakes include errors in computer coding. For research reproducibility, the entire research pipeline from formulating a research question to data acquisition, preparation, and analysis to dissemination of data and code should be designed to mitigate the potential for human bias and mistakes. Topics include reproducible programming (github, Rmarkdown), quality control techniques (code review), adjusting for multiple comparisons, honest characterization of uncertainty (confidence intervals, bootstrapping, and cross validation), and methods for power and sample size calculations.
Featured Speakers in the Department of Epidemiology & Biostatistics:
- Stathis Gennatas, MBBS, PhD, Assistant Professor - Reproducible Coding
- Chuck McCulloch, PhD, MS, MA, Professor - Power & Sample Size
- Aaron Wolfe Scheffler, PhD, Assistant Professor - Confidence Intervals, Bootstrapping, Cross Validation, & P-hacking
- Anusha Vable, ScD, MPH, Associate Professor - Code Review as a Simple Trick to Enhance Reproducibility, Accelerate Learning, and Improve the Quality of Your Team’s Research
Pre-requisite skills: For lectures, participants are expected to be familiar with basic statistical concepts (hypothesis testing, confidence intervals, multiple linear regression). For computer labs, participants are expected to execute R code and install packages
DATES (all sessions will take place 1:00pm-4:00pm Pacific Standard Time)
Monday April 15
Tuesday April 16
Wednesday April 17
PARTICIPATION FORMAT FOR THIS COURSE
ZOOM ONLY for this module.
COST
Although this module is free of charge, the same time, effort, and preparation contributes to the quality of each course. If you decide not to attend a class for which you have registered, the favor of your cancellation would be appreciated by e-mailing [email protected]
Program Directors
- June Maylin Chan, DSc, Professor & Vice Chair, Education, Epidemiology & Biostatistics
- Maria Glymour, ScD, MS, Professor, Epidemiology & Biostatistics
- Fei Jiang, PhD, MS, Assistant Professor, Epidemiology & Biostatistics
Funding Acknowledgement
National Institute of Aging/National Institutes of Health R25AG078149
Participation Formats
There are two possible participation formats, with some modules being offered hybrid (both Zoom and in-person). Please check the specific module for its participation format.
- In-person at UCSF Mission Hall: 550 16th street, San Francisco, CA 94158
- Online via Zoom