Data source from HERD Higher Edu Survey: https://www.nsf.gov/statistics/srvyherd/#tabs-2&toolsµ&profiles&tabs-2
- Determine useful comparisons for analysis and inform UA community clientel
- Determine Univ. Arizona Research Computing investment comparision to peer institutions
Bias localized to Federal Research Institutions for Federal NSF Dollars allocated to R&D where a positive correlation means closer ties
- H0: There is no bias in Federal Dollars allocated to R&D
- HA: Expectation: See bias toward increased expenditures for DoD/DoE et al affiliated Universities
Methods:
- Create geocoded plot of US with institutions and NSF funding received -1. Geocode data from https://nces.ed.gov/programs/edge/Geographic/SchoolLocations -- Used 2016-2017 Post-secondary data
- Merge list #25 from nsf website to the geocode data
- Use Heatmap for funding ranges seperated by arbitrary increments (likely in $100M)
- Geocodes are too tightly packed or use color glyph at zoom (Bokeh Datashader for region with glyphs)
- NEED to consider edge cases such as UA MED PHX or UA MED TUCSON are split
- Data files are split across years (1992 - 2016)