Social Media Usage by Demographic Groups

pew-social-media.R

library(tidyverse)
library(Kmisc)
## https://www.pewresearch.org/internet/fact-sheet/social-media/

# Read tables quickly ==========================================================
pew_01 <- read.table(
  text = "Facebook	Instagram	LinkedIn
Total	69%	40%	28%
Men	61%	36%	31%
Women	77%	44%	26%
Ages 18-29	70%	71%	30%
30-49	77%	48%	36%
50-64	73%	29%	33%
65+	50%	13%	11%
White	67%	35%	29%
Black	74%	49%	27%
Hispanic	72%	52%	19%
Less than $30K	70%	35%	12%
$30K-$49,999	76%	45%	21%
$50K-$74,999	61%	39%	21%
More than $75K	70%	47%	50%
High school or less	64%	30%	10%
Some college	71%	44%	28%
College graduate	73%	49%	51%
Urban	70%	45%	30%
Suburban	70%	41%	33%
Rural	67%	25%	15%",
  fill = TRUE,
  sep = "\t"
)

pew_02 <- read.table(
  text = "Twitter	Pinterest	Snapchat
Total	23%	31%	25%
Men	25%	16%	22%
Women	22%	46%	28%
Ages 18-29	42%	32%	65%
30-49	27%	34%	24%
50-64	18%	38%	12%
65+	7%	18%	2%
White	22%	34%	23%
Black	29%	35%	26%
Hispanic	23%	18%	31%
Less than $30K	12%	21%	25%
$30K-$49,999	29%	33%	27%
$50K-$74,999	22%	29%	29%
More than $75K	34%	40%	28%
High school or less	14%	22%	21%
Some college	26%	36%	32%
College graduate	33%	37%	23%
Urban	27%	30%	28%
Suburban	23%	32%	25%
Rural	18%	34%	18%
",
  fill = TRUE,
  sep = "\t"
)

pew_03 <- read.table(
  text = "YouTube	WhatsApp	Reddit
Total	81%	23%	18%
Men	82%	26%	23%
Women	80%	21%	12%
Ages 18-29	95%	24%	36%
30-49	91%	30%	22%
50-64	83%	23%	10%
65+	49%	10%	3%
White	79%	16%	17%
Black	84%	23%	17%
Hispanic	85%	46%	14%
Less than $30K	75%	23%	10%
$30K-$49,999	83%	20%	17%
$50K-$74,999	79%	19%	20%
More than $75K	90%	29%	26%
High school or less	70%	20%	9%
Some college	86%	16%	20%
College graduate	89%	33%	26%
Urban	84%	28%	18%
Suburban	81%	23%	21%
Rural	74%	9%	10%",
  fill = TRUE,
  sep = "\t"
)

pew_04 <- read.table(
  text = "TikTok	Nextdoor
Total	21%	13%
Men	17%	10%
Women	24%	16%
Ages 18-29	48%	5%
30-49	22%	17%
50-64	14%	16%
65+	4%	8%
White	18%	15%
Black	30%	10%
Hispanic	31%	8%
Less than $30K	22%	6%
$30K-$49,999	29%	11%
$50K-$74,999	20%	12%
More than $75K	20%	20%
High school or less	21%	4%
Some college	24%	12%
College graduate	19%	24%
Urban	24%	17%
Suburban	20%	14%
Rural	16%	2%",
  fill = TRUE,
  sep = "\t"
)

# Merge than get row names into column =========================================
pew <- bind_cols(pew_01, pew_02, pew_03, pew_04) %>%
  mutate(across(everything(), ~ as.numeric(gsub("%", "", .x)))) %>%
  rownames_to_column(var = "type") %>%
  filter(type != "Total") ## %>%
  ## select(-Nextdoor, -Snapchat, -Reddit) ## -WhatsApp, 

## Quickly
pew$category <- c(
  rep("Gender", 2), rep("Age", 4), rep("Race", 3), rep("Income", 4),
  rep("Education", 3), rep("Urbanism", 3)
)

pew <- pew %>%
  pivot_longer(
    !type & !category,
    names_to = "Platform", values_to = "Proportion"
  ) %>%
  mutate(
    Platform = factor(
      Platform,
      levels = c(
        "YouTube", "Facebook", "Instagram", "Pinterest", "LinkedIn",
        "Twitter", "TikTok", "WhatsApp", "Reddit", "Snapchat", "Nextdoor"
      )
    )
  )

# ggplot2 ======================================================================
p1 <- ggplot(pew %>% filter(category == "Gender")) +
  geom_col(
    aes(x = Platform, y = Proportion, colour = NULL, fill = Platform)
  ) +
  facet_wrap(~type, ncol = 2) +
  scale_fill_viridis_d()

pdf("pew_social_media_gender.pdf", width = 6, height = 3)
print(
  plot_nolegend(pdf_default(p1)) +
    theme(axis.text.x = element_text(angle = 45, hjust = 1))
)
dev.off()

p2 <- ggplot(pew %>% filter(category == "Age")) +
  geom_col(
    aes(x = Platform, y = Proportion, colour = NULL, fill = Platform)
  ) +
  facet_wrap(~type, ncol = 2) +
  scale_fill_viridis_d()

pdf("pew_social_media_age.pdf", width = 6, height = 6)
print(
  plot_nolegend(pdf_default(p2)) +
    theme(axis.text.x = element_text(angle = 45, hjust = 1))
)
dev.off()

p3 <- ggplot(pew %>% filter(category == "Race")) +
  geom_col(
    aes(x = Platform, y = Proportion, colour = NULL, fill = Platform)
  ) +
  facet_wrap(~type, ncol = 2) +
  scale_fill_viridis_d()

pdf("pew_social_media_race.pdf", width = 6, height = 6)
print(
  plot_nolegend(pdf_default(p3)) +
    theme(axis.text.x = element_text(angle = 45, hjust = 1))
)
dev.off()

p4 <- ggplot(pew %>% filter(category == "Income")) +
  geom_col(
    aes(x = Platform, y = Proportion, colour = NULL, fill = Platform)
  ) +
  facet_wrap(~type, ncol = 2) +
  scale_fill_viridis_d()

pdf("pew_social_media_income.pdf", width = 6, height = 6)
print(
  plot_nolegend(pdf_default(p4)) +
    theme(axis.text.x = element_text(angle = 45, hjust = 1))
)
dev.off()

p5 <- ggplot(pew %>% filter(category == "Education")) +
  geom_col(
    aes(x = Platform, y = Proportion, colour = NULL, fill = Platform)
  ) +
  facet_wrap(~type, ncol = 2) +
  scale_fill_viridis_d()

pdf("pew_social_media_education.pdf", width = 6, height = 6)
print(
  plot_nolegend(pdf_default(p5)) +
    theme(axis.text.x = element_text(angle = 45, hjust = 1))
)
dev.off()