Creating Technology for Social Change

Intro to Civic Media: Immigration Coverage with PageOneX


For my final project for Introduction to Civic Media with Professor Sasha Costanza-Chock, I have been working with the platform PageOneX in order to conduct a front page analysis of immigration coverage in major newspapers.

As mentioned in a previous blog, PageOneX is an online hosted platform that allows users to manually code front page newspaper coverage. Using scanned images that are pulled from, PageOneX allows users to manually highlight specific coverage that they are looking to examine. The interface for the platform allows you to first create “threads,” which are the titles for any specific set of newspapers that you wish to examine based on submitted topics. With my particular project, I will be looking at month-long threads across 3 major newspapers: New York Times, Los Angeles Times, and La Opinion. This means that a single month-long thread within my project will contain roughly 90 daily newspapers that must be manually coded. The coding process is as simple as clicking on the scanned image of a newspaper that is displayed, and then dragging a rendered rectangle to fit the area of the specific article. Once an area is selected, it can be resized or moved around as desired. One downside though, is that the image provided is very blurry and the only way to see it full sized is to constantly copy and paste the image’s link into a new tab. At the end of the coding process, the surface area of each coded article is calculated as a percentage and displayed as bars along an x/y axis.

The specific articles that I will be coding for this project are related to immigration. I have already been able to code the month of January 2010, which came out to 93 daily newspapers worth of coverage. Initially, I wanted to code a topic for coverage related to the DREAM Act. However, I soon discovered that I had to revise my method of coding for several reasons. First of all, I found that most articles in both English and Spanish languages talked about my selected subject using specific terms: immigrants, migrants, immigration. Most of the headings for coverage relating to immigration used those three terms. Also, I decided to include other topics that seemed related to immigration. These other topics included coverage of Arizona’s SB 1070 Bill, Sheriff Joe Arpaio, and immigrant detention. Perhaps I will go back and add other specific topic, such as “Arizona” or “SB 1070,” but for now they are all included under the topic “immigration.”

Out of the 93 daily newspapers that I was to scan, only 14 of them had articles related to immigration. The most simultaneous front page coverage that was given during the month of January 2010 came on the first day of the year, January 1, with over 11% combined coverage across the LA Times and La Opinion. The New York Times had none. However, since I only used a single color for the topic parameter for all three newspapers, the visualization provided combines all of the coverage instead of separating them. It would seem more useful to have a different color for each newspaper in order to create a visualization that shows which papers provided most coverage and when. Different colors will definitely be used in the future. Also, from the visualization provided by PageOneX, we can see that immigration coverage existed during 11 out of 31 days for all three papers combined, with the LA Times having 2/31 days, NY Times with 2/31 days, and La Opinion having a whopping 10/31 days. Before entering this analysis, I would have guess that Spanish-language media would have covered immigration more on their front pages, which turned out to be exactly true.

For the next rounds of coding sessions, I will be sure to add different colors for each newspaper in order to make the visualization easier to read. I may even have to re-code the month of January 2010 because the storage limitations on the free hosted site Heroku would not allow me to continue to code. However, Pablo Rey Mazón, the site’s creator, was kind enough to host an entire separate version of the PageOneX tool that will be specifically used for this project. Furthermore, I am still debating whether to include major events that were occurring during the months examined. For example, in January of 2010, the massive earthquake that some of us remember hit Haiti, and all of the newspapers that I was coding had several days worth of coverage related to the event.