CTSI Design Studio May 17 with Jim Chen: Heterogeneity in Data Analysis for Research

Published: March 8th, 2017

Category: Uncategorized

temp_design_studioDo you need assistance with the analysis of pilot data for your grant submission? Designing an experiment? Getting overwhelmed with a lot of data?

We can help!

Come to the CTSI Design Studio! Professional biostatisticians, epidemiologists and team members will consult on your project. Lunch will be provided!

The CTSI Biostatistics, Epidemiology, and Research Design (BERD) Program hosts these studios the third Wednesday of every month, free of charge. The studio is an opportunity for UF investigators to informally discuss their translational research ideas and questions with BERD faculty members. By providing free, informal and early access to BERD experts, the studio aims to help investigators save time and strengthen their study designs and results.

Any UF investigator is invited to attend to discuss new study proposals, ideas for data analysis, or a statistical or data analysis topic of interest. Each studio typically begins with a topical presentation (30 minutes including Q&A), followed by small group discussions focused on attendees’ specific study designs. Studio attendance is limited to ensure participants benefit from individual time with BERD experts.

BERD also provides basic, walk-in statistical guidance for research on Mondays, Wednesdays and Fridays from 1 p.m. to 2 p.m. in Clinical and Translational Research Building (CTRB) 5212. In addition, you may schedule an appointment with a BERD Data Analyst or an epidemiologist.

Speaker: Jim Chen, M.D., Ph.D.
Topic: Heterogeneity in Data Analysis for Research
When: May 17, 2017 (third Wednesday of the month), 11:30 a.m. to 1 p.m.
Where: CTRB 2161 South

Heterogeneity has not been well considered in research. Ignored heterogeneity can lead to wrong conclusions. Heterogeneity can be due to the study population itself and can also arise through research design (i.e., cluster sampling and group randomization). Three approaches are available to deal with heterogeneity: Stratified analysis, analysis treating heterogeneity as it is (e.g., latent class analysis, group-based developmental trajectory analysis) and analysis controlling for heterogeneity (e.g., mixed effect modeling). These methods allow researchers to draw solid conclusions when used properly in data analysis.

Lunch will be served. Limited to 8 participants. For questions, contact April Braxton at abraxton@ufl.edu.

RSVP for May 17, 2017 Design Studio

 

Keep an eye out for the next studio, which will be announced in this space as well as in the CTSI email newsletter.