Abstract: Across biomedical research areas, scientists make and use code to facilitate data collection, generate data through computational simulations, investigate patterns in data through descriptive statistics and visualizations, and run statistical analyses to connect data to hypotheses. This talk will give an overview of an ongoing research program about the present and future of scientific software with an emphasis on applications to biomedical research. First, we will review findings from a recently-published study of how scientists adopt new software tools for data-intensive research, which sheds light on the pressures currently facing scientists who wish to expand their data analysis toolkit. Second, we will discuss new research that will look at which researchers are using generative AI tools for programming, how researchers reason about the validity of generated code, and how scientific practices and teamwork may be changing as a result.  

Bio:  Elle O'Brien, Ph.D. is a lecturer and research investigator at the University of Michigan. Dr. O'Brien trained at the University of Washington in computational neuroscience methods with a particular focus on the auditory nervous system. Through this work, she became interested in broad questions about how scientists develop software to collect, analyze, and share data and statistical models. Currently, Dr. O'Brien is researching how scientists navigate the use of generative AI tools for programming.

This year’s SHRP schedule is available here: https://purdue.edu/TPAN/hearing/shrp_schedule

Titles and abstracts of all SHRP talks are here: https://purdue.edu/TPAN/hearing/shrp_abstracts