[Add hero image or embedded video here.]

Understanding access to internet and its usage in Mexico

Team Members: William Huang, Cuauhtemoc (Temo) Ojeda, Ee Suk (Isaac) Ahn, Jared Moore

Visualizations: Final Project | Initial Prototype

Project Goals

Our topic for our project will be “Access of information in the states in Mexico.” Throughout this project, we will answer questions like “How are different subsets of different population reachable in Mexico?” “How does different groups in each distribution relate to use of information?” “How does each state differ in access of information?” and more other. Initially, we will get data for people in Mexico, and the data will include census data (age, population, income, etc.), technology usage, and more. We will display the data and sort by information accessibility, then we want to exemplify the biggest differences we found from people in different location. To do so, we start by visualizing people in different location with different indicators, and then contrast it with those from other places. We will show maps and different graphs that could show the differences we have found. Our interactive visualization could be showing more detailed information when states are hovered over, switching datasets using interactivity tools, zooming, and others.

Motivation

We (residents of the U.S.) likely take access to the internet for granted. Some do not have such privilege. Digital access determines a person’s ability to serve as a full member of society. The first step in improving access is determining those who donot have access.

The Insituto Nacional de Estadística y Geografía (INEGI) began conducting special surveys on digital technologies in 2015 with the National Survey on Availability and Use of Information Technologies in Households (ENDUTIH). They believe that “it is essential to have statistics that are accurate, timely and with the widest possible geographical breakdown on” the development of access to information technology.

Their work “aims to obtain information on the availability and use of information and communication technologies in households and their use by individuals aged 6 years and over in Mexico to generate statistical information on the subject and support the decision-making on public policy issues.” It includes “103,000 housing units, distributed in 32 states and 49 selected cities” and was conducted from May 6 to July 29, 2016.

Our Contribution

INEGI’s work is great, but it is not enough. Their graphs come straight from excel and do little to present information in a perceptually easy or exploratory manner. On a first read of their report, a reader might not even realize the breadth of the data or understand that she can examine free from responses. We aim to build on the work of INEGI, making their data more easily accessible and understandable through an interactive web visualization.

At the very least we hope to inform members the UW CSE community about the unfortunate impact location has on a person’s access to technology. We not only want our audience to realize that Chiapas has significantly less access to internet than Baja California Sur, but we also want them to step back and think: access to internet is mediocre in almost all of Mexico.

Throughout this project, we have attempted to answer questions such as “How are different subsets of different population reachable in Mexico?” “How do different groups in each distribution relate to use of information?” “How does each state differ in access of information?,” etc.

Approach

Maps serve as a natural encoding of location based data. We have data about Mexican's access to internet at a state granularity for location. Thus we chose to display the states of Mexico as a Mercator projection with each colored by a binned proportion of internet access [link]. Because Mexico is near the equator and not too large a Mercator projection works well. We rely on the research from colorbrewer to determine the change in color each of the five steps and chose green because we feel it qualitatively represents connectivity.

Our visualization allows for comparison between states [link] and drill down on individual states. We decided to show the type of internet per state and comparisons between states with bar charts [link] [link]because they allow our audience to quickly view the most common type of internet connection in each state and make comparison between states perceptually easy. On hover of a state, we display the name and percentage of the state's population connected to the internet. [link]

While we could only aggregate the survey data to the state level, we had access to individual survey responses – people. This was powerful. Drawing on the narrative approaches for visualization, we chose draw a story out of the data and highlight two respondents and their families. Juan [link] and Maria [link] illustrate the class and location differences in access.

Similarly, we curated a number of common responses to the question “Why don’t you have internet in your household?” [link] These were telling: at times humorous, weird, and tragic. We hope that these responses also allow for greater empathy on the part of our audience and elucidate the people behind aggregations.

In our pursuit of accessibility, we also implemented a Spanish version [link] of our webpage and hope to allow Mexicans to gain insight from our page as well.

Future Work

The INEGI dataset is huge – we touched a very small part of it and were only able to display and even smaller part. Our works merely sets an example for how government research can be more easily displayed to a wider audience; we would like to see more of this.

We briefly explored information on the respondent level of the survey, but we wish that we could have shared more. Allowing viewers to explore and compare respondents and aggregate based on respondent features would make this visualization much more powerful and alive.

We primarily focused on aggregations and survey questions directly related to internet, but he data has more to offer. A more comprehensive analysis would examine correlations between access to electricity, city size, or level of education and access to internet. We could spend an entire year on just this dataset.

Data

TODO: fill in citation Report: http://www.inegi.org.mx/saladeprensa/aproposito/2017/internet2017_Nal.pdf Methods: http://internet.contenidos.inegi.org.mx/contenidos/Productos/prod_serv/contenidos/espanol/bvinegi/productos/nueva_estruc/702825091057.pdf