Biostatistics Support

The Biostatistics Office contributes to the NINDS mission by providing objective, high-quality statistical solutions through proper implementation of statistical methods and by promoting the use of rigorous quantitative methods.
The scope of the work may include:

  • Providing statistical expertise and advice to investigators at NINDS;
  • Engaging in collaborative research with NINDS investigators;
  • Providing education on statistical methods and various analytical techniques through lecture series and/or seminars;
  • Mentoring students and fellows who are pursuing a career in biomedical or biostatistical research;
  • Conducting research in statistical methodology related to NINDS research;
  • Participating in Protocol Development Meetings (PDM) at CTU providing statistical input on the study protocols;
  • Participating in Data Management Meetings for recently approved protocols to oversee the quality and management of key outcome measures and other analytical endpoints; and
  • Serving on the NINDS Scientific Review Committee (SRC), intramural Institutional Review Boards (IRB), and attending advisory committee meetings.
info-circle
For statistical assistance
Please contact our office’s shared mailbox: biostatistics@ninds.nih.gov or Gina Norato to have your question properly routed.

Biostatistics Office Staff
Gina Norato, ScM – Office Lead, Statistician
Tianxia Wu, PhD – Statistician
Henry Roberts, PhD – Statistician
Ken Cheung, PhD – Contract Statistician
 

Tips & Tools

video
Statistical Education
The Biostatistics Office prioritizes meaningful and accessible education for clinical trainees and investigators. Regular seminar series and educational lectures are conducted throughout the year. View recent recordings of the 2022 Statistics Lecture Series on NIH Videocast.

Statistical Tip of the Month

Bootstrapping is a statistical approach that can be used to describe inference of a sample about a population. We can use this to create simulated datasets with characteristics similar to our original dataset for the purposes of estimation or describing variability. This procedure can be done by sampling “with replacement” from the original dataset, to create many (perhaps thousands) of new simulated datasets of the same size n. From each sample, we can extract some summary statistic, such as a mean, to get an understanding of the possible distribution of means we can get from similar datasets.