Water: Everyone Needs It, but Who Has It?

Overview and Motivation

Water is a fundamental human resource, yet a significant portion of the world today lacks regular access to clean drinking water. Many factors - wealth, rainfall, geographic location, to name a few - could be contributing to this disparity, but how and how much? With our visualization, we intend to illustrate the correlation and magnitude of that correlation between these factors and clean drinking water access.

Related Work

Here is a list of works we referenced while researching our topic, and some thoughts on them.

• Although not incredibly informative, this visualization is easy to digest and highlights the main issues with water access.
• This graphic is closer to what we are going for. There is a bit of information-overload, but it's quite informative and aesthetic!
• This site is rich with drinking water visualzations. At this point we are beginning to get a firmer idea of what we want to do, but it's really interesting to poke around and see the different factors this author thought to highlight in terms of drinking water. Really interesting!

Questions

• Economic Indicators (like GDP)
• Latitude
• Tempurature
• Population Density / Population
• Conflicts / War
• Rainfall

Sketches

Data

Our inspiration
Water and sanitation

Supplementary data by country

• GDP
• Poverty
• Rainfall
• Conflict
• Latitude/Longitude
• Population density

Exploratory Data Analysis

We have a number of distinct datasets, so our plan is to go through them one-by-one and compile them into a single data structure that we can easily perform analysis on. Roughly, this will be organized as a multi-dimensional map, where the first dimension represents the year the data is from, the second dimension represents the country the data is from, the third dimension represents the category of data, the fourth dimension represents the data label, and finally the last dimension represents the underlying value of the data point. For example, structured_data[2000][China][drinking_water][at_least_basic_water_access] == 80, representing 80% of China's population.

Unfortunately, our poverty dataset is abysmally sparse & we can't find decent alternatives online. We will rely on GDP for the time being to represent a country's economic success. Here's an image of the poverty dataset (note: each column should be full!):

Another concession: our precipitation data is missing a few years we are considering otherwise. We will use linear interpolation to fill in the gaps - we aren't expecting large changes in precipitation over just a couple of years.

Wrangling the data was a far more intensive process than expected. The general process was to begin by either loading each CSV into JavaScript and logging the data, or opening the CSV/spreadsheet in Excel to clean first. One of the websites, WorldBank, had a bug where one could only download Excel files, not CSVs, which contributed to the time it took to get the data ready. However, while we did end up replacing some of the datasets for an inability to download them or simply not having enough data, we were able to accomplish our goal of fiting the data into a shape that's easy to manipulate. This will make prototyping our visualization and improving upon it much easier. We've found & refined datasets for GDP, population density, regional conflicts, latitude, precipitation, average tempurature, and water access, which we would say is more than enough for some meaningful analysis and visualization.

Design Evolution

After the Tuesday, 12/1 lecture, we got the impression that maps are a generic and overdone method of visualization. Thus, we've cut the map portion of our visualization and will instead be focusing on improving our existing views.

Implementation

After cleaning the data, we focused on making a working prototype for the scatterplot and latitude bar chart. We ran into some trouble at first passing functions asynchronously, but it was solved by returning a Promise object from cleanData and chaining .then()s afterward. With that small hiccup out of the was a simple matter to pack the cleaned data into the appropriate format for the visualization we required.

Evaluation

The user studies gave us some good insight on the strengths of our project. Things commonly given props include our layout and the customizability of our data. It also exposed flaws and bugs. Notably, users wanted more labels on axes so you don't have to comb through the paragraph explanations to make sense of the graphs. Users noted a text overlap issue on the water percentage bar when changing the year to 2017. They requested wider bars, bigger dots and more responsive tooltips to make exploring the visualization easier. They suggested trimming outliers that made no sense, like countries with 0 GDP. They suggested keyword highlighting to make the paragraphs easier to follow. They also suggested a legend for the color scale.

Design Re-Evolution

After our feedback from milestone 2 and peer evaluation, we ultimately decided that a map would be a beneficial addition to the visualization after all. When discussing data from each country around the world, users just expect a map. A map doesn't convey as much real information as some of our other graphics, but it makes sense that it can be a helpful supplement to the graphs and allows for a new way to explore the data. Thus, we are adding the map back ahead of the presentation.

Milestone 1

Most of our work before milestone 1 was behind the scenes. We spent a lot of time, thinking, planning/sketching, investigating & collecting data, and cleaning data! However, we did mannage to have some of our visualization up-and-running for this milestone. Our basic scatterplot and latitude chart are up, though they aren't pretty! We have a lot of work to do.

Milestone 2

We made a lot of visual progress since milestone 1! We now have basic tooltips, three graphics, and a layout that looks more like a dashboard than an html page. We discussed for a while whether a map would be beneficial or simply detract from the rest of the visualization and decided to let the data presentation speak for itself without the map.

Presentation

We spent a lot of time thinking about how to best explain our visualization - we didn't just want to make a tech demo or a day-by-day re-visiting of creating the visualization. Instead, we decided that the most interesting & valuable presentation would discuss the fundamental questions we had about our problem, and therefore how our visualization could be used to answer these questions. We knew we only had a few minutes to present, so we'd have to be efficient with all of the time we're given, which is why we are recording the presentation in advance.

Project Journal

Throughout our project, we reserved this process book for major updates, and kept a more detailed journal which we entered into (almost) every time we worked on the project. Listed below are the entries from this journal.

Thursday, 11/19

Monday, 11/23

Thursday, 12/3

Saturday, 12/5

Monday, 12/7

Tuesday, 12/8

Gameplan

Notes

Saturday, 12/12

Sunday, 12/13

Monday, 12/14

Monday, 12/14