Pageviews analysis

Kaitlin Kavlie, PSYC 541

Assignment 2: Page Views Analysis

1. I created a graph of views of the Gun control article on Wikipedia over a period of a few years. The Las Vegas shooting which took place on October 1st of 2017 and the Parkland, Florida shooting which took place February 14th of 2018 are the points of focus when examining the data.

I used this first code below to create a data set called ‘gun’ of views of the Wikipedia article ‘Gun control’ ranging from 1/1/2017 until 1/1/2019.

gun <- article_pageviews(article = "Gun_control", start = as.Date("2017-1-1"), end = as.Date("2019-1-1"))

glimpse(gun)

Rows: 731
Columns: 8
$ project     <chr> "wikipedia", "wikipedia", "wikipedia", "wikipedia", "wikipedia", "wikipedia", "wikipedia", ~
$ language    <chr> "en", "en", "en", "en", "en", "en", "en", "en", "en", "en", "en", "en", "en", "en", "en", "~
$ article     <chr> "Gun_control", "Gun_control", "Gun_control", "Gun_control", "Gun_control", "Gun_control", "~
$ access      <chr> "all-access", "all-access", "all-access", "all-access", "all-access", "all-access", "all-ac~
$ agent       <chr> "all-agents", "all-agents", "all-agents", "all-agents", "all-agents", "all-agents", "all-ag~
$ granularity <chr> "daily", "daily", "daily", "daily", "daily", "daily", "daily", "daily", "daily", "daily", "~
$ date        <dttm> 2017-01-01, 2017-01-02, 2017-01-03, 2017-01-04, 2017-01-05, 2017-01-06, 2017-01-07, 2017-0~
$ views       <dbl> 285, 393, 434, 544, 635, 620, 431, 464, 610, 595, 686, 688, 591, 313, 396, 441, 674, 660, 7~

Then I used the following code to create a graph of the ‘gun’ data set.

gun %>% 
  ggplot(aes(x = date, y = views)) +
  geom_line(color = "pink") +
  labs(x= "Date", y = "Wikipedia Views", title = "Views of Wikipedia's Gun Control Article")

As shown from the graph created the views of Wikipedia’s Gun Control article peaked right around the time of both the Las Vegas and the Florida shooting.

2. I created a table showing the top 10 days for viewing the ‘Gun control’ article on Wikipedia using the code below.

gun %>% 
  select(article, views, date) %>%
  top_n(10, views)%>%
  arrange(-views)%>%
  datatable(class = 'cell-border stripe') %>% 
  formatStyle("date", backgroundColor = "plum")%>%
  formatStyle("views",backgroundColor = "lightpink")

NA

As you can see from the analysis some of the top 10 dates for highest views on Wikipedia’s Gun control article were within a week of the Las Vegas shooting, as well as within a week of the Florida shooting.

3. Using the codes shown below, I created a table of the top 10 viewed articles the day after the Las Vegas shooting, as well as a table for the day after the Florida shooting.

top_articles(start = as.Date("2017-10-2")) %>%
  select(article, views) %>%
  filter(!article == "Main_Page", !article == "Special:Search")

top_articles(start = as.Date("2018-2-15")) %>%
  select(article, views) %>%
  filter(!article == "Main_Page", !article == "Special:Search")

As you can see from the analyses above, the day after both shootings the top Wikipedia articles viewed were topics related to shootings. The Las Vegas shooting occurred on 10/1/2017, the day after one of the top viewed articles was ‘2017 Las Vegas Strip Shooting’. The Florida shooting occurred on 2/14/2018, the day after some of the top viewed articles were ‘Columbine High School massacre’ and ‘School shootings in the United States’.

4. Using a ggplot I compared views of the Gun Control article on Wikipedia 1 week before and 2 weeks after the Vegas and Florida shootings.

First, I used the codes below to create a data set called ‘LV’ for views of the Wikipedia ‘Gun control’ article 1 week before and 2 weeks after the Las Vegas shooting. The mutate coding altered the graph so the time shown is labeled starting 1 week prior to and 2 weeks after the shooting. Then the ggplot coding actually graphed the information.

LV <- article_pageviews(article = "Gun_control", start = as.Date("2017-9-24"), end = as.Date("2017-10-15"))

LV <- LV %>% 
  mutate(day = -7:14) %>% 
  mutate(event = "Las Vegas")

LV %>% 
  ggplot(aes(x = day, y = views)) +
  geom_line()

Then, I used the codes to create a data set called ‘FL’ for views of the Wikipedia ‘Gun control’ article 1 week before and 2 weeks after the Florida shooting. The mutate coding then altered the graph’s time label to start 1 week prior to and 2 weeks after the shooting. Then the ggplot coding graphed the data set.

FL <- article_pageviews(article = "Gun_control", start = as.Date("2018-2-7"), end = as.Date("2018-2-28")) 

FL <- FL %>% 
  mutate(day = -7:14) %>% 
  mutate(event = "Florida")

FL %>% 
  ggplot(aes(x = day, y = views)) +
  geom_line()

Lastly, I created a data set called ‘Shootings’ using the bind rows code to combine the ‘LV’ and ‘FL’ data sets. Then the ggplot coding shown below graphed the combined data, allowing for them to be visually compared to each other.

Shootings <- bind_rows(LV, FL)

Shootings %>% 
  ggplot(aes(x = day, y = views, color = event)) +
  geom_line() +
  theme_minimal() +
  labs(x = "Days before/after Shooting", 
       y = "Wikipedia Views", 
       color = "Event", 
       title = "Views of the Wikipedia Gun Control Article before and after Two Mass Shootings")

As shown by the analyses run above the views of Wikipedia’s ‘Gun control’ article were the lowest the week before both of the shootings. The views of the article peaked 1-2 days after each shooting. After the views peaked a few days after each shooting, the views tapered off but still remained higher 2 weeks after each shooting than 1 week before.