Overcoming Fears (My Own) Teaching Reproducible Research, Big Data and Data Mining in Nursing and Public Health Education

• Activity Number: 74 - Challenges and Approaches to Teaching Statistics in the Health Sciences
• Type: Contributed
• Date/Time: Sunday, July 29, 2018 : 4:00 PM to 5:50 PM
• Sponsor: Section on Teaching of Statistics in the Health Sciences
• Author(s): Melinda Higgins*
• Companies: Emory University
• Keywords: reproducible; git; rstudio; github; education; data mining

Abstract

To capitalize on the explosion of health data, big data computing platforms and data mining are critical for nursing and public health scientists. To address these needs, in spring 2017 we implemented our first course on "Big Data Analytics for Healthcare" (with second cohort in spring 2018). This presentation will cover lessons learned from both instructor and student perspectives. Statistical modeling and data mining were taught with R and RStudio with Git version control and Github. AWS computing was also introduced. Reproducible research principles and workflow were stressed. I expected more technical issues and student fears which were unfounded, exceeding both my and student expectations. Final student projects were challenging and well executed. Several student exemplars included: microbiome data analysis from the American Gut Project; integration of microclimate sensors and macroclimate regional weather data to improve heat risk warnings for agricultural workers; and analysis of social-media parenting blogs using web-scraping and textual data mining. Each of these exemplars addressed one or more of the social, behavioral, economic, or environmental determinants of health.