Descriptive analysis

Basic Data Information

library(tidyverse)
library(lubridate)

neda_nine <- read_rds(here::here(
  "data",
  "NEDA_week_nineteen.rds"
)) %>%
  mutate(data_year = 2019) %>%
  select(-media) %>%
  filter(created_at_tweet %within% interval(ymd("2019-02-25"), ymd("2019-03-04"))) %>%
  mutate(interval_created = int_length(interval(ymd("2019-02-25"), created_at_tweet)))

neda_twenty <- read_rds(here::here(
  "data",
  "NEDA_week_twenty.rds"
)) %>%
  mutate(data_year = 2020) %>%
  filter(created_at_tweet %within% interval("2020-02-24", ymd("2020-03-02"))) %>%
  mutate(interval_created = int_length(interval(ymd("2020-02-24"), created_at_tweet)))


neda <- bind_rows(neda_nine, neda_twenty)

Below are the basic statistics (total number of tweets, number of unique users and date):

• Number of total tweets:

library(gt)

total_tweets <- neda %>%
  group_by(data_year) %>%
  count() %>%
  ungroup()

total_tweets

## # A tibble: 2 x 2
##   data_year     n
##       <dbl> <int>
## 1      2019 19432
## 2      2020 16463

• Number of unique users:

unique_users <- neda %>%
  group_by(data_year) %>%
  distinct(name) %>%
  count() %>%
  ungroup()

unique_users

## # A tibble: 2 x 2
##   data_year     n
##       <dbl> <int>
## 1      2019 10773
## 2      2020 10402

• Date from:

2019 NEDA week officialy from 25th to March 3st : 2019-02-25 00:00:03, 2019-03-03 22:00:26

2020 NEDA week from February 24th to March 1st: 2020-02-24 00:02:16, 2020-03-01 22:59:04

Who tweets?

neda %>%
  group_by(data_year) %>%
  count(screen_name, sort = T) %>%
  slice(1:10) %>%
  gt() %>%
  tab_options(
    heading.title.font.size = 13,
    column_labels.font.size = 11,
    table.font.size = 11
  ) %>%
  tab_header(
    title = "# of different tweets by user"
  ) %>%
  fmt_number(
    columns = vars(n),
    decimals = 0
  )

# of different tweets by user
screen_name	n
2019
NEDAstaff	375
Dianev44	256
zacharydmallory	224
Grace_Durbin	111
SupermodelEmme	102
KidsMatterOK	77
EDFColorado	69
heavenlysins66	56
JenniSchaefer	56
TheProjectHEAL	56
2020
NEDAstaff	114
uofleatlab	74
CheriLev	66
NaopakSasvim	34
EDCareDenver	33
MentalHealthMil	31
YazzyLaMer	31
EDFColorado	28
KelseyEClark	28
drmuhlheim	26

Accounts with the highest number of users?

neda %>%
  group_by(data_year) %>%
  select(screen_name, followers_count) %>%
  arrange(desc(followers_count)) %>%
  distinct(screen_name, .keep_all = T) %>%
  slice(1:20) %>%
  gt() %>%
  tab_options(
    heading.title.font.size = 13,
    column_labels.font.size = 11,
    table.font.size = 11
  ) %>%
  tab_header(
    title = "Firts 20th users by # of followers"
  ) %>%
  fmt_number(
    columns = vars(followers_count),
    decimals = 0
  )

Firts 20th users by # of followers
screen_name	followers_count
2019
instagram	36,665,765
MTV	15,499,665
MTVNEWS	5,160,025
WomensHealthMag	4,581,386
MensHealthMag	4,516,375
TeenVogue	3,340,148
inquirerdotnet	2,792,707
Ginger_Zee	2,340,517
Pinterest	2,337,657
Jimparedes	1,751,509
harpersbazaarus	1,677,033
seventeen	1,359,212
NIMHgov	1,153,600
womenshealth	936,617
HRC	811,703
HHSGov	754,150
dosomething	750,146
ABC7NY	653,570
teddyboylocsin	646,794
Allure_magazine	576,799
2020
instagram	35,993,421
gmanews	5,533,600
parentsmagazine	4,763,121
CNNnews18	4,342,366
GMA	3,707,198
inquirerdotnet	2,938,922
sadierob	1,871,043
DZMMTeleRadyo	1,332,968
NIMHgov	1,163,247
womenshealth	932,087
NYDailyNews	736,888
AlvaroAlvaradoC	652,868
SELFmagazine	507,126
nutribullet	442,739
Xiaxue	374,580
MentalHealthAm	333,155
TrevorProject	290,826
WCVB	287,301
TWLOHA	278,281
raphablueberry	254,924

Which tweets were retweeted the most?

neda %>%
  filter(str_detect(text, "^RT @NEDAstaff")) %>%
  group_by(data_year) %>%
  count(retweeted_status_id, sort = TRUE) %>%
  left_join(neda %>% select(id_tweet, text, created_at_tweet),
    by = c("retweeted_status_id" = "id_tweet")
  ) %>%
  slice(1:10) %>%
  select(-retweeted_status_id, created_at_tweet) %>%
  gt()

n	text	created_at_tweet
2019
176	Today kicks of #NEDAwareness Week 2019 from Feb 25-Mar 3. Our theme, "Come as You Are", highlights NEDA’s movement… https://t.co/cFJhr1BqJJ	2019-02-25 17:30:07
98	For #NEDAwareness Week, @Instagram and @NEDA have teamed up to let individuals at all stages of body acceptance and… https://t.co/exGloFska1	2019-02-26 17:09:04
92	Thank you to the @EmpireStateBldg for lighting up in blue and green for #NEDAwareness Week! https://t.co/p2SbhcDbMI	2019-02-27 02:24:37
56	#NEDAwareness Week kicks off this Monday and buildings across the country will light up in blue and green to shine… https://t.co/9bscDufL3A	2019-02-25 00:38:00
53	#NEDAwareness Week is Here! https://t.co/GgJEcJOWyB	2019-02-25 17:03:04
45	This #NEDAwareness Week (2/25-3/4), take and share our screening tool - you never know who might need it! https://t.co/dQ0iCIC9L7.	2019-02-25 22:08:00
41	For #NEDAwareness Week, NEDA Ambassadors Ryan Sheldon @BingeConfession and Karla Mosley @karlamosley share how Eat… https://t.co/VZiY65soDw	2019-02-28 16:16:01
36	If you are in a crisis, text 'NEDA' to 741741 to connect with a trained volunteer from Crisis Text Line. #NEDAwareness	2019-02-26 20:59:02
31	#NEDAwareness Week kicks off this Monday and buildings across the country will light up in blue and green to shine… https://t.co/KZMSmrTps1	2019-02-26 04:13:00
22	This #NEDAwareness Week (2/25-3/4), take and share our screening tool - you never know who might need it! https://t.co/dQ0iCIC9L7.	2019-02-27 02:23:00
2020
125	Happy #NEDAwareness Week 2020! 🎉 We encourage everyone to embrace this year’s theme, #ComeAsYouAre: Hindsight is 20… https://t.co/NPU8E1cHNO	2020-02-24 11:41:01
48	Thank you to the @EmpireStateBldg for lighting up in NEDA’s signature blue and green colors tonight! We are proud t… https://t.co/fnnLMtUYRU	2020-02-26 02:56:27
36	In honor of #NEDAwareness Week, #NEDAmbassador, Gina Susanna (@nourishandeat), shares strategies via @i_weigh on ho… https://t.co/D978UGBu3Q	2020-02-26 23:15:02
27	#DYK: 1 in 3 people with an #eatingdisorder identify as male? On the NEDA blog, @DrStuartBMurray and his @USC colle… https://t.co/ygg9zpvEKj	2020-02-25 21:20:03
26	This #NEDAwareness Week, we want to hear about YOUR experience with #eatingdisorders, body image, and/or exercise i… https://t.co/0tgtZTNHJz	2020-02-24 18:27:02
21	NEDA and @instagram teamed up this #NEDAwareness week to encourage you to #ComeAsYouAre and use the social media pl… https://t.co/5a155uZmv3	2020-02-28 16:01:07
20	We love seeing all of the #NEDAwareness Week reflections! To minimize triggers and harmful content, please take a l… https://t.co/YyamTcFxVT	2020-02-24 17:04:01
18	Throughout #NEDAwareness week, monuments and buildings across America will be lighting up in NEDA’s signature blue… https://t.co/jq3VfGCDuP	2020-02-24 21:07:02
17	#NEDAwareness Week starts tomorrow!! Is your school or organization helping to spread awareness about food, body im… https://t.co/vdRaG2yNBB	2020-02-24 00:14:11
17	Come as You Are and join the NEDAwareness Week conversation on social media. Don’t forget to use the official campa… https://t.co/vesASwtlSi	2020-02-24 13:58:01

Tweets per hour during neda

library(scales)

neda %>%
  select(
    created_at_tweet, screen_name, retweeted_status_user_screen_name, text,
    interval_created, data_year
  ) %>%
  mutate(rec_screen_name = case_when(
    screen_name != "NEDAstaff" ~ "Others users",
    TRUE ~ "NEDAstaff"
  )) %>%
  mutate(round_created_at_tweet = round_date(created_at_tweet, unit = "hour")) %>%
  count(data_year, round_created_at_tweet, rec_screen_name) %>%
  ggplot(aes(x = round_created_at_tweet, y = n, colour = rec_screen_name)) +
  geom_line() +
  labs(
    title = "",
    colour = "Tweets posted by:",
    x = "NEDAwareness week",
    y = "# of tweets"
  ) +
  scale_x_datetime() +
  facet_wrap(data_year ~ ., ncol = 1, scales = "free_x") +
  theme_light() +
  theme(legend.position = "bottom") +
  scale_color_brewer(palette = "Set1")

Retweets and likes of the content posted by users from other accounts

neda_20_hist <- read_rds(here::here("data", "NEDA_historical_twenty.rds")) %>%
  # filter(screen_name == "NEDAstaff") %>%
  select(id_tweet, created_at_tweet, retweet_count, favorite_count) %>%
  mutate(data_year = 2020) %>%
  arrange(desc(created_at_tweet)) %>%
  group_by(id_tweet) %>%
  slice(1) %>%
  ungroup()

neda_19_hist <- read_rds(here::here("data", "NEDA_historical_nineteen.rds")) %>%
  # filter(screen_name == "NEDAstaff") %>%
  select(id_tweet, created_at_tweet, retweet_count, favorite_count) %>%
  mutate(data_year = 2019) %>%
  arrange(desc(created_at_tweet)) %>%
  group_by(id_tweet) %>%
  slice(1) %>%
  ungroup()

d <- bind_rows(neda_19_hist, neda_20_hist)

d %>%
  mutate(id_tweet = as.character(id_tweet)) %>%
  select(id_tweet, retweet_count, favorite_count, data_year) %>%
  pivot_longer(
    cols = -c(data_year, id_tweet), values_to = "values",
    names_to = "variables"
  ) %>%
  group_by(data_year, variables) %>%
  count(values, name = "number_of_tweets") %>%
  ungroup() %>%
  filter(values > 0) %>%
  mutate(variables = factor(variables,
    level = c(
      "retweet_count",
      "favorite_count"
    ),
    labels = c(
      "Retweets",
      "Likes"
    )
  )) -> dd

dd %>%
  mutate(data_year = as_factor(data_year)) %>%
  ggplot(aes(x = values, y = number_of_tweets, color = data_year)) +
  geom_point() +
  scale_y_log10(breaks = log_breaks()) +
  scale_x_log10(
    breaks = trans_breaks("log10", function(x) 10^x),
    labels = trans_format("log10", math_format(10^.x))
  ) +
  scale_color_brewer(palette = "Set1") +
  theme_light() +
  xlab("# rt/likes from a user in collection") +
  ylab("# Tweets") +
  facet_wrap(. ~ variables)

Gender and LIWC

2019 data:

read_rds(here::here("data", "neda_liwc_gender_nineteen.rds")) %>%
  select(gender, id) %>%
  distinct() %>%
  count(gender, sort = T)

## # A tibble: 3 x 2
##   gender     n
##   <chr>  <int>
## 1 f        762
## 2 u        671
## 3 m        313

2020 data:

read_rds(here::here("data", "neda_liwc_gender_twenty.rds")) %>%
  select(gender, id) %>%
  distinct() %>%
  count(gender, sort = T)

## # A tibble: 3 x 2
##   gender     n
##   <chr>  <int>
## 1 u        193
## 2 f        173
## 3 m         65