In our simple linear regression example, we looked at the correlation between a college’s distribution of majors and their patents and patent citations, and then we tried to predict the “innovativeness” of a college based on the percentage of STEM undergraduate majors. This gave us some insight into how we might be able to predict a college’s innovativeness based on the percentage of STEM majors--using the number of patents as an approximation for innovation--however, outcomes usually depend on more than one variable.
Now, we’ll create a more robust multiple linear regression analysis to look at what other variables might predict any college’s number of patents--and innovativeness--such as a school’s endowment, dollars spent on undergraduate students, and upward social mobility.
- Output work (Excel data sets from in-class examples) used to look at colleges’ role in upward social mobility
- Innovation Rates by College, Opportunity Insights (also here + data dictionary)
- College Level Characteristics from the IPEDS Database and the College Scorecard - mrc_table10; Opportunity Insights
- Microsoft Excel + ToolPak](https://support.office.com/en-us/article/load-the-analysis-toolpak-in-excel-6a63e598-cd6d-42e3-9317-6b40ba1a66b4)