Skip to main content
. 2020 May 13;6(5):e03956. doi: 10.1016/j.heliyon.2020.e03956

Figure 3.

Figure 3

[Top panel]: Model fit visualization representing observed SAT decile (x-axis) versus fitted SAT percentiles (y-axis) for private institutions (left; beta regression pseudo R-squared=0.58) and public institutions (right; beta regression pseudo R-squared=0.40). There were no private institutions reporting average SAT scores in the bottom decile. [Bottom panel]: Density plots of the impact of a 5% redistribution into majors with positive (left) or negative (right) model coefficients in Table 1.