ursing <- read.csv("WiscNursingHome.csv")
nursing <- nursing[complete.cases(nursing), ]
names(nursing)
# [1] "hospID"    "CRYEAR"    "TPY"
# [4] "NUMBED"    "SQRFOOT" "MSA"
# [7] "URBAN"    "PRO"       "TAXEXEMPT"
# [10] "SELFFUNDINS" "MCERT"    "ORGSTR"
# [13] "LOGTPY"
nursing$LOGTPY <- log(nursing$TPY)

# a)
str(nursing$PRO)
str(nursing$TAXEXEMPT)

# Using three binary vairables creates the perfect collinearity,
# which is not allowed in the regression model.

# d)
fit <- lm(LOGTPY ~ URBAN + PRO + TAXEXEMPT + SELFFUNDINS + MCERT,
      data = nursing)
summary(fit)
# Call:
# lm(formula = LOGTPY ~ URBAN + PRO + TAXEXEMPT + SELFFUNDINS +
#    MCERT, data = nursing)
#
# Residuals:
#    Min    1Q Median    3Q    Max
# -1.94492 -0.30261 0.01761 0.31195 1.63007
#
# Coefficients:
# Estimate Std. Error t value Pr(>|t|)
# (Intercept) 4.13338 0.08044 51.382 < 2e-16 ***
# URBAN    0.19136 0.03623 5.282 1.71e-07 ***
# PRO       -0.34184 0.06022 -5.677 2.01e-08 ***
# TAXEXEMPT -0.21551 0.06240 -3.453 0.000587 ***
# SELFFUNDINS 0.12452 0.03743 3.327 0.000925 ***
# MCERT    0.35080 0.06272 5.593 3.20e-08 **

Problem Set 5 Problem 1. This problem uses data from the Preventi...