Recoding Variables Final Paper
Chelsy Rodriguez
4/28/2021
I am interested to know how full-time and part-time respondents (employ_2019) differ their attitude towards their annual income. I will be comparing them based on how satisfied they are with their job (satisf_Job_2018), how satisifed they are with their annual income (satisf_Income_2018), and if they agree or disagree about the government having responsibility to reduce the income differences between people with lower income and people with higher income (diff_inc_2019), and their feeling towards welfare recipients (wr_2019).
In the above example, the employ_2019 variable will break up my respondents into two groups (full-time and part-time employees), and the other four variables represents the behavior/attitudes which I will be investigating as I compare full-time employees and part-time employees.
library(readr)
library(dplyr)##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(ggplot2)## Warning in file(con, "r"): cannot open file '/var/db/timezone/zoneinfo/
## +VERSION': No such file or directorylibrary(knitr)
Voter2019 <- read_csv("/Users/chelsyrodriguez/Downloads/Voter Data 2019.csv")##
## ── Column specification ────────────────────────────────────────────────────────
## cols(
## .default = col_double(),
## weight_18_24_2018 = col_logical(),
## izip_2019 = col_character(),
## housevote_other_2019 = col_character(),
## senatevote_other_2019 = col_character(),
## senatevote2_other_2019 = col_character(),
## SenCand1Name_2019 = col_character(),
## SenCand1Party_2019 = col_character(),
## SenCand2Name_2019 = col_character(),
## SenCand2Party_2019 = col_character(),
## SenCand3Name_2019 = col_character(),
## SenCand3Party_2019 = col_character(),
## SenCand1Name2_2019 = col_character(),
## SenCand1Party2_2019 = col_character(),
## SenCand2Name2_2019 = col_character(),
## SenCand2Party2_2019 = col_character(),
## SenCand3Name2_2019 = col_character(),
## SenCand3Party2_2019 = col_character(),
## governorvote_other_2019 = col_character(),
## GovCand1Name_2019 = col_character(),
## GovCand1Party_2019 = col_character()
## # ... with 108 more columns
## )
## ℹ Use `spec()` for the full column specifications.## Warning: 800 parsing failures.
## row col expected actual file
## 2033 weight_18_24_2018 1/0/T/F/TRUE/FALSE .917710168467982 '/Users/chelsyrodriguez/Downloads/Voter Data 2019.csv'
## 2828 weight_18_24_2018 1/0/T/F/TRUE/FALSE 1.41022291345592 '/Users/chelsyrodriguez/Downloads/Voter Data 2019.csv'
## 4511 weight_18_24_2018 1/0/T/F/TRUE/FALSE 1.77501243840922 '/Users/chelsyrodriguez/Downloads/Voter Data 2019.csv'
## 7264 weight_18_24_2018 1/0/T/F/TRUE/FALSE 1.29486870319614 '/Users/chelsyrodriguez/Downloads/Voter Data 2019.csv'
## 7277 weight_18_24_2018 1/0/T/F/TRUE/FALSE 1.44972719707603 '/Users/chelsyrodriguez/Downloads/Voter Data 2019.csv'
## .... ................. .................. ................ ......................................................
## See problems(...) for more details.VoterData2019<-Voter2019 %>%
mutate(Employmentstatus = ifelse(employ_2019==1,"Full-time",
ifelse(employ_2019==2,"Part-time","NA")),
SatisfiedlifeJob = ifelse(satisf_Job_2018==1,"VS",
ifelse(satisf_Job_2018==2,"SS",
ifelse(satisf_Job_2018==3,"NS nor DS",
ifelse(satisf_Job_2018==4,"SD","NA")))),
SatisfiedlifeIncome = ifelse(satisf_Income_2018==1,"VS",
ifelse(satisf_Income_2018==2,"SS",
ifelse(satisf_Income_2018==3,"NS nor DS",
ifelse(satisf_Income_2018==4,"SD",
ifelse(satisf_Income_2018==5,"VD","NA"))))),
DifferenceIncome = ifelse(diff_inc_2019==1,"Strongly agree",
ifelse(diff_inc_2019==2,"Somewhat agree",
ifelse(diff_inc_2019==3,"Neither agree or disagree",
ifelse(diff_inc_2019==4,"Somewhat disagree",
ifelse(diff_inc_2019==5,"Strongly disagree",
ifelse(diff_inc_2019==8,"skipped","NA")))))),
FeelingAboutWelfareRecipients = ifelse(wr_2019>100,NA,wr_2019))%>%
select(Employmentstatus,SatisfiedlifeJob,SatisfiedlifeIncome,DifferenceIncome,FeelingAboutWelfareRecipients)%>%
filter(Employmentstatus %in% c("Full-time","Part-time"))
head(VoterData2019)## # A tibble: 6 x 5
## Employmentstatus SatisfiedlifeJob SatisfiedlifeIn… DifferenceIncome
## <chr> <chr> <chr> <chr>
## 1 Full-time <NA> <NA> Strongly agree
## 2 Full-time VS NS nor DS Strongly disagr…
## 3 Full-time <NA> <NA> Somewhat agree
## 4 Part-time NS nor DS VD Strongly agree
## 5 Full-time VS VS Somewhat agree
## 6 Part-time NS nor DS SD Strongly agree
## # … with 1 more variable: FeelingAboutWelfareRecipients <dbl>Categorical Variable #1
Crosstab
table(VoterData2019$SatisfiedlifeJob,VoterData2019$Employmentstatus) %>%
prop.table(2)##
## Full-time Part-time
## NA 0.07261641 0.08897485
## NS nor DS 0.15410200 0.23597679
## SD 0.10421286 0.08317215
## SS 0.39356984 0.36363636
## VS 0.27549889 0.22823985round(2)## [1] 2Visualization: Stacked chart
VoterData2019 %>%
group_by(SatisfiedlifeJob,Employmentstatus) %>%
summarize(n=n()) %>%
mutate(percent=n/sum(n)) %>%
ggplot()+
geom_col(aes(x=SatisfiedlifeJob,y=percent,fill=Employmentstatus))## `summarise()` has grouped output by 'SatisfiedlifeJob'. You can override using the `.groups` argument.
The stacked chart for variable 1 shows us that full-time employees have a higher percentage than part-time employees on how satisfied or dissatisfied they are with their job. There are more full-time employees that are somewhat dissatisfied (SD) with their job than part-time employees. As for part-time employees, it shows they’re neither satisfied (NS) or dissatisfied (DS) with their job.
Statistical Test: Chi-square test
chisq.test(VoterData2019$SatisfiedlifeJob,VoterData2019$Employmentstatus)##
## Pearson's Chi-squared test
##
## data: VoterData2019$SatisfiedlifeJob and VoterData2019$Employmentstatus
## X-squared = 23.145, df = 4, p-value = 0.0001184The p-value indicates the scientific notation as 0.0001184 which is a smaller value then 0.05. Therefore there is a statistically significant relationship on how full-time and part-time employees differ their attitudes in how satisfied they are with their job.
Categorical Variable #2
Crosstab
table(VoterData2019$SatisfiedlifeIncome,VoterData2019$Employmentstatus) %>%
prop.table(2)##
## Full-time Part-time
## NA 0.01385809 0.01547389
## NS nor DS 0.16186253 0.19148936
## SD 0.17960089 0.23210832
## SS 0.41851441 0.31528046
## VD 0.08647450 0.13926499
## VS 0.13968958 0.10638298round(2)## [1] 2Visualization: Stacked Chart
VoterData2019 %>%
group_by(SatisfiedlifeIncome,Employmentstatus) %>%
summarize(n=n()) %>%
mutate(percent=n/sum(n)) %>%
ggplot()+
geom_col(aes(x=SatisfiedlifeIncome,y=percent,fill=Employmentstatus))## `summarise()` has grouped output by 'SatisfiedlifeIncome'. You can override using the `.groups` argument.
The stacked chart for variable 2 shows the same higher percentage in full-time employees then part-time employees on how satisfied or dissatisfied they are with their income. Full-time employees are very satisfied (VS) with their income. As for part-time employees, they are very dissatisfied (VD) with their income.
Statistical Test: Chi-square test
chisq.test(VoterData2019$SatisfiedlifeIncome,VoterData2019$Employmentstatus)##
## Pearson's Chi-squared test
##
## data: VoterData2019$SatisfiedlifeIncome and VoterData2019$Employmentstatus
## X-squared = 33.557, df = 5, p-value = 2.917e-06The p-value shows the scientific notation as 2.917e-06 which is a larger value than .05.It shows a strong evidence for the null hypothesis. There isn’t a statistically significant relationship on how full-time and part-time employees differ their attitudes on how satisfied they are with their income.
Categorical Variable 3
Table Comparing means
VoterData2019%>%
group_by(Employmentstatus)%>%
summarize(Avg_FeelingAboutWelfareRecipients = mean(FeelingAboutWelfareRecipients, na.rm=TRUE))## # A tibble: 2 x 2
## Employmentstatus Avg_FeelingAboutWelfareRecipients
## * <chr> <dbl>
## 1 Full-time 48.8
## 2 Part-time 51.1Visualization: Bar Chart comparing means
VoterData2019 %>%
group_by(FeelingAboutWelfareRecipients,Employmentstatus) %>%
summarize(n=n()) %>%
mutate(percent=n/sum(n)) %>%
ggplot()+
geom_col(aes(x=Employmentstatus,y=percent,fill=Employmentstatus))## `summarise()` has grouped output by 'FeelingAboutWelfareRecipients'. You can override using the `.groups` argument.
Full-time employees has a higher percentage on feelings towards welfare recipients compared to part-time employees.
Visualization: Histogram of feelings about welfare recipients based on employment status
VoterData2019 %>%
filter(Employmentstatus %in% c("Full-time","Part-time")) %>%
ggplot()+
geom_histogram(aes(x=FeelingAboutWelfareRecipients))+
facet_wrap(~Employmentstatus)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.## Warning: Removed 2797 rows containing non-finite values (stat_bin).
Comparison of sample means of feeling about welfare recipients
Fulltime_data <- VoterData2019 %>%
filter(Employmentstatus=="Full-time")
sample(Fulltime_data$FeelingAboutWelfareRecipients,40) %>%
mean(na.rm=TRUE)## [1] 43.33333Fulltime_data <- VoterData2019 %>%
filter(Employmentstatus=="Full-time")
sample(Fulltime_data$FeelingAboutWelfareRecipients,40) %>%
mean()## [1] NAreplicate(10000,sample(Fulltime_data$FeelingAboutWelfareRecipients,40)%>%mean(na.rm=TRUE)) %>%
data.frame() %>%
rename("mean"=1)%>%
ggplot()+
geom_histogram(aes(x=mean),fill="red")## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.## Warning: Removed 1 rows containing non-finite values (stat_bin).
Parttime_data <- VoterData2019 %>%
filter(Employmentstatus=="Part-time")
sample(Parttime_data$FeelingAboutWelfareRecipients,40) %>%
mean(na.rm=TRUE)## [1] 60.8Parttime_data <- VoterData2019 %>%
filter(Employmentstatus=="Part-time")
sample(Parttime_data$FeelingAboutWelfareRecipients,40) %>%
mean()## [1] NAreplicate(10000,sample(Parttime_data$FeelingAboutWelfareRecipients,40)%>%mean(na.rm=TRUE)) %>%
data.frame() %>%
rename("mean"=1)%>%
ggplot()+
geom_histogram(aes(x=mean),fill="blue")## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.## Warning: Removed 8 rows containing non-finite values (stat_bin).
T-test
Newdata <- VoterData2019 %>%
filter(Employmentstatus %in% c("Full-time","Part-time"))
t.test(FeelingAboutWelfareRecipients~Employmentstatus, data = Newdata)##
## Welch Two Sample t-test
##
## data: FeelingAboutWelfareRecipients by Employmentstatus
## t = -0.80625, df = 152.08, p-value = 0.4214
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -7.986139 3.357118
## sample estimates:
## mean in group Full-time mean in group Part-time
## 48.75153 51.06604The p-value shows that there isn’t a statistically significant difference between full-time and part-time respondents that differ their feelings towards welfare recipients.
Conclusion
According to my analysis, most full-time employers are somewhat dissatisfied with their job however they are very satisfied with their income. Most part-time employers didn’t indicate neither they’re satisfied nor dissatisfied but they are very dissatisfied with their income. Afterall, it is accurate that part time employees would have that response since they are working under 40 hours a week. As for how they feel about welfare recipients, both full-time and part-time employers almost feel the same.