ADV 200 Week12

gss_cat <- read.csv("/Users/eunseokim/Desktop/gss_cat.csv", stringsAsFactors = TRUE)

Explanation

Converting character strings to factors is beneficial because many variables are categorical in this dataset, such as marital, race, and income.

table(gss_cat$marital)

## 
##      Divorced       Married Never married     No answer     Separated 
##          3383         10117          5416            17           743 
##       Widowed 
##          1807

table(gss_cat$race)

## 
## Black Other White 
##  3129  1959 16395

missing_counts <- sapply(gss_cat, function(x) sum(is.na(x)))

print(missing_counts)

##       X    year marital     age    race rincome partyid   relig   denom tvhours 
##       0       0       0      76       0       0       0       0       0   10146

names(which(missing_counts > 0))

## [1] "age"     "tvhours"

gss_cat$tvhours <- ifelse(is.na(gss_cat$tvhours),
mean(gss_cat$tvhours, na.rm = TRUE),
gss_cat$tvhours)

Explanation

Replacing missing values with the mean is logical for tvhours, as it’s a numeric variable representing hours.

5. Discussion about replacing missing values for all variables

For numeric variables like tvhours, replacing missing values with the mean is feasible.

For categorical variables like marital and race, replacing with the mode might be more appropriate.