knitr::opts_chunk$set(echo = TRUE)

Introduction

Add text here

Initially Proposed Model

Total expenditures: Add more text here

mod_totexp <- lm(totexp/1e6 ~ msda + phdfld + ggacpe + initcirc + type, 
                 data = arl_totexp)
summary(mod_totexp)
## 
## Call:
## lm(formula = totexp/1e+06 ~ msda + phdfld + ggacpe + initcirc + 
##     type, data = arl_totexp)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -44.633  -4.644  -0.066   3.929  43.233 
## 
## Coefficients:
##               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)  7.150e+00  3.489e+00   2.050  0.04325 *  
## msda         2.048e-03  9.928e-04   2.063  0.04197 *  
## phdfld       1.726e-01  4.244e-02   4.067  0.00010 ***
## ggacpe       1.112e-04  6.579e-05   1.690  0.09433 .  
## initcirc     7.316e-05  9.479e-06   7.718 1.38e-11 ***
## typeS       -8.359e+00  2.632e+00  -3.176  0.00203 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 10.6 on 92 degrees of freedom
## Multiple R-squared:  0.6722, Adjusted R-squared:  0.6544 
## F-statistic: 37.74 on 5 and 92 DF,  p-value: < 2.2e-16
pred_totexp <- predict(mod_totexp, newdata = uva_totexp)
compare(pred_totexp, uva_totexp$totexp/1e6)
## Predicted: $27.78 
## Actual:    $40.03
# also try msda in place of gradstu

Materials Expenditures: add more text here

mod_explm <- lm(explm/1e6 ~ msda + phdfld + ggacpe + initcirc + type, 
                 data = arl_explm)
summary(mod_explm)
## 
## Call:
## lm(formula = explm/1e+06 ~ msda + phdfld + ggacpe + initcirc + 
##     type, data = arl_explm)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -15.6722  -1.8072  -0.4022   1.4197  22.0490 
## 
## Coefficients:
##               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)  5.722e+00  1.604e+00   3.567 0.000576 ***
## msda         7.519e-04  4.566e-04   1.647 0.103004    
## phdfld       8.095e-02  1.952e-02   4.147 7.48e-05 ***
## ggacpe       4.069e-05  3.026e-05   1.345 0.181966    
## initcirc     2.525e-05  4.359e-06   5.794 9.54e-08 ***
## typeS       -4.876e+00  1.210e+00  -4.029 0.000115 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 4.874 on 92 degrees of freedom
## Multiple R-squared:  0.6186, Adjusted R-squared:  0.5979 
## F-statistic: 29.85 on 5 and 92 DF,  p-value: < 2.2e-16
pred_explm <- predict(mod_explm, newdata = uva_explm)
compare(pred_totexp, uva_explm$explm/1e6)
## Predicted: $27.78 
## Actual:    $14.03
# also try msda in place of gradstu