Methods SIG - Leadership/Facilitator Contact Info

¡Bienvenidos!

Welcome to the Methods SIG,

 We are excited about your interest in joining our group. We discuss possible ideas for statistical methods that are relevant to the analysis of data from HCHS/SOL in our calls. As you are aware, HCHS/SOL used a complex survey sampling design for the selection of the participants and it is a longitudinal study with the second in-person visit ongoing from October 2014 to September 2017. The participants are also interviewed by phone annually. This provides a wealth of data, but also can cause technical hitches in the analysis. It is important to have appropriate statistical analysis tools for various complications. In anticipation of potential complications in analyzing the longitudinal data and disease incidence data from the complex survey design, the Methods SIG meets on a monthly basis to discuss emerging problems and potential solutions.

Examples of issues discussed are: How to handle missing covariates data for longitudinal analysis? What if the covariates are missing at random (MAR)? What if they are missing not at random (NMAR)? How do we handle covariates which are measured with error? What if the magnitude of the error is related to the underlying measure itself? Is the specification of higher level clustering needed for longitudinal analysis? What is the influence of the sampling weights on the association studies? How do we design cost- and resource- efficient studies for future ancillary studies?

We encourage HCHS/SOL biostatisticians or HCHS/SOL affiliated biostatisticians who are interested in investigating new statistical methodology to join the Methods SIG. We also encourage HCHS/SOL or HCHS/SOL affiliated investigators from all disciplines who have unusual statistical problems to join the Methods SIG. The Methods SIG meetings will provide a forum for discussion of all issues and topics related to statistical methods, such as: 1) Discussion of complications for analysis; 2) Formulating statistical problems; 3) Potential direction and solution for the statistical problems; 4) Manuscript ideas; 5) Presentations of findings; 6) Ancillary study proposal ideas. It’s a great forum for early and mid-career investigators to discuss their statistical methods manuscript proposals and ideas.

Our regular recurring day/time for the Methods SIG meetings is (to be determined).

We look forward to your joining the Methods SIG and our monthly calls. New ideas are welcome!

Jianwen Cai, PhD
Cary C. Boshamer Distinguished Professor
Vice Chair Biostatistics, UNC Gillings School of Global Public Health
Principal Investigator, HCHS/SOL Coordinating Center