Chapter 15 - Factors

Load the libraries needed for these exercises.

library(tidyverse)
library(forcats)

15.3 - General Social Survey

Problem 1

Explore the distribution of rincome (reported income). What makes the default bar chart hard to understand? How could you improve the plot?

There are so many bars that the x-axis labels overlap.

ggplot(gss_cat, aes(rincome)) +
  geom_bar()

Let’s rotate the labels.

ggplot(gss_cat, aes(rincome)) +
  geom_bar() +
  theme(axis.text.x = element_text(angle=90))

Problem 2

What is the most common relig in this survey? What’s the most common partyid?

gss_cat %>%
  count(relig) %>%
  top_n(1)

## # A tibble: 1 x 2
##   relig          n
##   <fct>      <int>
## 1 Protestant 10846

gss_cat %>%
  count(partyid) %>%
  top_n(1)

## # A tibble: 1 x 2
##   partyid         n
##   <fct>       <int>
## 1 Independent  4119

Problem 3

Which relig does denom apply to? How can you find out without a table? How can you find out with a vizualization?

denom applies to Protestant.

gss_cat %>%
  group_by(denom) %>%
  count(relig)

## # A tibble: 47 x 3
## # Groups:   denom [30]
##    denom           relig          n
##    <fct>           <fct>      <int>
##  1 No answer       No answer     93
##  2 No answer       Christian      2
##  3 No answer       Protestant    22
##  4 Don't know      Christian     11
##  5 Don't know      Protestant    41
##  6 No denomination Christian    452
##  7 No denomination Other          7
##  8 No denomination Protestant  1224
##  9 Other           Protestant  2534
## 10 Episcopal       Protestant   397
## # ... with 37 more rows

gss_cat %>%
  group_by(relig, denom) %>%
  summarize(n = n()) %>%
  ggplot(aes(relig, denom, size = n)) +
    geom_point() +
    theme(axis.text.x=element_text(angle = 90))

15.4 - Modifying factor order

Problem 1

There are some conspicuously high numbers in tvhours. Is the mean a good summary?

Exteme values bias the mean. tvhours is skewed to the right and median may be a better summary statistic.

gss_cat %>%
  ggplot(aes(tvhours)) +
    geom_histogram()

mean(gss_cat$tvhours, na.rm = TRUE)

## [1] 2.980771

median(gss_cat$tvhours, na.rm = TRUE)

## [1] 2

Problem 2

For each factor in gss_cat identify whether the order of the levels is arbitrary or principled.

marital: arbitrary
age: principled
race: arbitrary
rincome: principled
partyid: arbitrary
relig: arbitrary
denom: arbitrary

Problem 3

Why did moving “Not Applicable” to the front of the levels move it to the bottom of the plot?

Factors are shown in reverse order from top to bottom when coord_flip is used.

gss_cat %>%
  mutate(race = fct_relevel(race, "Not applicable")) %>%
  ggplot(aes(race)) +
    geom_bar() +
    coord_flip() +
    scale_x_discrete(drop = FALSE)

15.5 - Modifying factor levels

Problem 1

How have the proportions of people identifying as Democrat, Republican, and Independent changed over time?

gss_cat %>%
  mutate(partyid = fct_collapse(partyid,
    other = c("No answer", "Don't know", "Other party"),
    rep = c("Strong republican", "Not str republican"),
    ind = c("Ind,near rep", "Independent", "Ind,near dem"), 
    dem = c("Not str democrat", "Strong democrat"))) %>%
  filter(partyid != "other") %>%
  ggplot(aes(x = year, ..prop.., fill = partyid)) +
    geom_bar(position = "dodge") +
    scale_x_continuous(breaks = c(2000, 2002, 2004, 2006, 2008, 2010, 2012, 2014))

Problem 2

How could you collapse rincome into a small set of categories?

gss_cat %>%
  mutate(rincome = fct_collapse(rincome, 
    `More than $10,000` = c("$10000 - 14999", "$15000 - 19999", "$20000 - 24999", "$25000 or more"),
    `Less than $10,000` = c("Lt $1000", "$1000 to 2999", "$3000 to 3999", "$4000 to 4999", "$5000 to 5999", "$6000 to 6999", "$7000 to 7999", "$8000 to 9999"))) %>%
  mutate(rincome = fct_lump(rincome, n = 3)) %>%
  count(rincome, sort = TRUE)

## # A tibble: 4 x 2
##   rincome               n
##   <fct>             <int>
## 1 More than $10,000 10862
## 2 Not applicable     7043
## 3 Less than $10,000  2153
## 4 Other              1425