PBA Extra Credits
Evangeline Olivia 109006206
2022-11-29
Opening Libraries
require(dplyr)
require(knitr)
require(pollster)
require(kableExtra)Question 1
setwd("/Users/olivia/Documents/Documents/Study/Semester 5/PBA")
judges <- read.csv(file = 'judges.csv')
temp<-xtabs(~woman+republican,judges)
rownames(temp)<-c("Men","Woman")
colnames(temp)<-c("Democrat","Republican")
temp## republican
## woman Democrat Republican
## Men 82 124
## Woman 27 11gender_rep_table<-temp/sum(temp);
rownames(gender_rep_table)<-c("Men","Woman")
colnames(gender_rep_table)<-c("Democrat","Republican")
gender_rep_table## republican
## woman Democrat Republican
## Men 0.33606557 0.50819672
## Woman 0.11065574 0.045081971. How many judges are in this data set?
cat("Sum of Judges:",sum(temp))## Sum of Judges: 2442. What proportion of the judges are men?
cat("Men Proportion:",round(sum(gender_rep_table[1,])*100,3),"%")## Men Proportion: 84.426 %3. Is the party composition different for male and female judges?
Yes it is different, we can see from the table below.| Democrat | Republican | |
|---|---|---|
| Men | 82 | 124 |
| Woman | 27 | 11 |
| Democrat | Republican | |
|---|---|---|
| Men | 0.3360656 | 0.5081967 |
| Woman | 0.1106557 | 0.0450820 |
Question 2
Histogram of progressive_vote
hist(judges$progressive_vote,main = "Histogram of progressive_vote",breaks = 10,col="lightcoral")
mean(judges$progressive_vote)## [1] 0.4351555median(judges$progressive_vote)## [1] 0.422619The histogram is more similar to a right-skewed histogram. As a supporting information I have provided the mean and median.
It shows that the median is more left rather than the mean which is the same as the characteristic of a right-skewed graph.
mean(judges$progressive_vote)## [1] 0.4351555median(judges$progressive_vote)## [1] 0.422619Where is the region of highest density of this variable?
The highest density is in between 0.25 and 0.50
Question 3
Create a new factor variable called judges$gender_party
- “F_Demo” for female judges appointed by Democratic presidents.
- “F_Repub” for female judges appointed by Republican presidents.
- “M_Demo” for male judges appointed by Democratic presidents.
- “M_Repub” for male judges appointed by Republican presidents.
judges$gender_party<-case_when(judges$woman==1 & judges$republican==0 ~ "F_Demo",
judges$woman==1 & judges$republican==1 ~ "F_Repub",
judges$woman==0 & judges$republican==0 ~ "M_Demo",
judges$woman==0 & judges$republican==1 ~ "M_Repub")Use tapply() to calculate the mean of progressive_vote in each of these groups and store this vector as gender_party_means.
gender_party_means<-tapply(judges$progressive_vote,judges$gender_party,mean)
gender_party_means## F_Demo F_Repub M_Demo M_Repub
## 0.4547162 0.3069867 0.5062359 0.3952614Plot these means using a barplot
barplot(height = gender_party_means,col="lightcoral")
Does anything stand out to you?
- Based on the bar plot we can see that both female and male democratic have a higher proportion on the progressive_vote
- Based on the party association the democrat have a higher proportion the progressive_vote
- Based on the gender we can see that Male who votes on republican or democratic have higher proportion on the progressive_vote than Woman
Question 4
Create a variable called judges$any_girls that is 1 when the judge has at least 1 girl and 0 otherwise.
judges$any_girls<-if_else(judges$girls<1|is.na(judges$girls)==TRUE,0,1)
head(judges)## name child circuit girls progressive_vote race religion
## 1 Alarcon, Arthur L. 3 9 1 0.0000000 3 4
## 2 Aldisert, Ruggero 3 3 1 0.6666667 1 4
## 3 Aldrich, Bailey 2 1 0 0.3333333 1 1
## 4 Alito, Samuel A., Jr. 2 3 1 0.5000000 1 4
## 5 Altimari, Frank X. 4 2 1 0.5000000 1 4
## 6 Ambro, Thomas NA 3 NA 0.0000000 NA NA
## republican sons woman yearb gender_party any_girls
## 1 0 2 0 1925 M_Demo 1
## 2 0 2 0 1919 M_Demo 1
## 3 1 2 0 1907 M_Repub 0
## 4 1 1 0 1950 M_Repub 1
## 5 1 3 0 1928 M_Repub 1
## 6 0 NA 0 NA M_Demo 0Create a subset of the data called parents that contains judges that have at least one child.
parents<- subset(judges,child>=1)
head(parents)## name child circuit girls progressive_vote race religion
## 1 Alarcon, Arthur L. 3 9 1 0.0000000 3 4
## 2 Aldisert, Ruggero 3 3 1 0.6666667 1 4
## 3 Aldrich, Bailey 2 1 0 0.3333333 1 1
## 4 Alito, Samuel A., Jr. 2 3 1 0.5000000 1 4
## 5 Altimari, Frank X. 4 2 1 0.5000000 1 4
## 7 Anderson, R. Lanier, III 3 5 NA 0.2500000 1 2
## republican sons woman yearb gender_party any_girls
## 1 0 2 0 1925 M_Demo 1
## 2 0 2 0 1919 M_Demo 1
## 3 1 2 0 1907 M_Repub 0
## 4 1 1 0 1950 M_Repub 1
## 5 1 3 0 1928 M_Repub 1
## 7 0 NA 0 1936 M_Demo 0Create an object called ATE that is the difference in means of progressive_vote between judges that have at least one girl versus those that have no girls among those judges with any children.
ATE <- tapply(parents$progressive_vote, parents$any_girls, mean)
ATE## 0 1
## 0.3976401 0.4518024ATE<-ATE[2]-ATE[1]
cat("Difference : ",ATE)## Difference : 0.05416231Can we interpret the result causally (i.e., can we safely say that the difference, if there is any, is caused by the judges having a daughter)?
No, we cannot interpret the result by just seeing how many daugther the judges have can influence the decision since there are many other factors that can influence someone’s decision. We can also see the difference is so small (0.05…) so we cannot interpret just like that.