If you have chosen to work with Excel, please run above three models and complete the following tables.

Model $1$: Run a regression model and use being a member of network and find out its impact on hospital cost? $($Model $1)$ Model $2$: For the $2$nd model run a regression model and use being a member of network and find out its impact on hospital cost and hospital revenue? $($Model $2)$ Model $3$: For the $3$rd model run a regression model and use being a member of network and find out its impact on ratio$-$Medicare$-$discharge and ratio$-$Medicaid$-$discharge.

Based on your finding please recommend $3$ policies and discuss the impact of being on a network on hospital cost, hospital revenue and out its impact on ratio$-$Medicare$-$discharge and ratio$-$Medicaid$-$discharge. Do you recommend keeping membership for a hospital? Why or why not? # You can remove the # from the sink command to keep all log in txt$.$

# sink$("$C:$/$UMUC$/$week$5$exercise.txt$")$

# Step $1$: Install package dplyr & load it

# install.packages$(\text{'}$dplyr$\text{'})$

library$($dplyr$)$

# Step $2$: Read your data

# Pl change the location of file

hosp $<-$ read.csv$("$C:$/$UMUC$/$HMGT$400$HOSPITAL.csv$",$ header$=$T$,$ sep $=\text{'},\text{'})$

#Step $3$: See the variables' names

names $($hosp$)$

#step $4$ : generate benefit variable

hosp$benefit $<-($hosp$total$\_$hosp$\_$revenue$-$hosp$total$\_$hosp$\_$cost$)$

hosp$medicare$\_$discharge$\_$ratio $<-($hosp$total$\_$hospital$\_$medicare$\_$discharg$/$hosp$total$\_$hospital$\_$discharges$)*100$

hosp$medicaid$\_$discharge$\_$ratio $<-($hosp$total$\_$hospital$\_$medicaid$\_$discharg$/$hosp$total$\_$hospital$\_$discharges$)*100$

#step $5$ : generate own$1$ as a categorical variable

## Ownership

## $0)$ non$-$for$-$profit

## $1)$ for profit

## $2)$ Public

## $3)$ Other

own$1<-$ factor$($hosp$own, levels $=$ c$(0,1,2,3))$

#step $6$ : Run logit models

# Model $1$

model$1<-$ glm$($system$\_$member ~ own$1+$ hospital$\_$beds,family$=$binomial$($link$=$'logit'$),$data$=$hosp$)$

summary$($model$1)$

# Model $2$

model$2<-$ glm$($system$\_$member ~ own$1+$ hospital$\_$beds $+$ total$\_$hosp$\_$revenue,family$=$binomial$($link$=$'logit'$),$data$=$hosp$)$

summary$($model$2)$

# Model $3$

model$3<-$ glm$($system$\_$member ~ own$1+$ hospital$\_$beds $+$ total$\_$hosp$\_$revenue $+$ medicare$\_$discharge$\_$ratio $+$ medicaid$\_$discharge$\_$ratio,family$=$binomial$($link$=$'logit'$),$data$=$hosp$)$

summary$($model$3)$

# sink$()$

model$1<-$ glm$($system$\_$member ~ hospital$\_$beds,family$=$binomial$($link$=$'logit'$),$data$=$hosp$)$

summary$($model$1)$

model$2<-$ glm$($system$\_$member ~ hospital$\_$beds $+$ total$\_$hosp$\_$revenue,family$=$binomial$($link$=$'logit'$),$data$=$hosp$)$

summary$($model$2)$

model$3<-$ glm$($system$\_$member ~ hospital$\_$beds $+$ total$\_$hosp$\_$revenue $+$ medicare$\_$discharge$\_$ratio $+$ medicaid$\_$discharge$\_$ratio,family$=$binomial$($link$=$'logit'$),$data$=$hosp$)$

summary$($model$3)$