Data Visualizations - Background
For over a decade Vote Smart worked at various methods of building the Voter’s Self-Defense System. During this process Vote Smart received a great many very flattering commendations (see What they are Saying). However, the focus was always on completing the system, and completing the system meant having compiled detailed applications of employment on thousands of candidates whether they cooperated or not.
We thought we had accomplished this goal in 2002 and made plans to officially announce our success, but setbacks delayed this.
Congressional Quarterly pulled the plug on our Key Votes Division. We rebuilt it. Then it became apparent that the major political parties had effectively dismantled our ability to acquire useful issue positions by convincing candidates not to provide those issue positions in a meaningful way for fear of opposition research. So we researched and provided their positions for them.
In 2008 we finally had the entire system up and operating and spent close to $1 million in full-page newspaper ads and a 45,000 mile national tour to deliver the Voter’s Self-Defense System to the country. What better way, we thought, to conquer misinformation and tortured truth than our immense database of abundant, accurate, factual information on those seeking to represent us. It would now be up to citizens to use it, to flock to it. But they didn’t.
Although heavily used by political scientists, journalists, political activists and opposition researcher we were only able to reach 12.5% of the American public. That left 87.5% unaware of Vote Smart and its Voter’s Self-Defense program. It was as if Vote Smart had gone on a quest for the Holy Grail, found it, spent years bringing it home, only to find out no one gave a damn.
We asked ourselves why? And then we began researching the Voter’s Self-Defense System user’s experience. We discovered that time was a serious issue. On average it took a citizen using the system 40 minutes to look up a single candidate on a single issue and find all the data we had compiled on it in each of our six categories. The information was there but citizens would need to spend hours looking up all of the candidates they had to decide on and all the issues they were interested in. It was too much data, too deep, too rich.
The Voter’s Self-Defense System had to be simplified, lead users into one quick, tasty, juicy bite. Then maybe they could learn to enjoy and digest the reality that the Voter’s Self-Defense System represented.
In 2010 Vote Smart tested VoteEasy, an interactive tool that allowed citizens to match their own issue positions with those of the candidates on a small sampling of twelve issues. Users type in their zip code, answer a few questions on key issues and watch as the candidates’ campaign yard signs advance or recede depending on the level of their agreement. If the users question the movement of a sign or want more detail they simply click on it and see the candidate’s public record on that issue.
Our modest test of VoteEasy went viral. So many people used it that we were unable to handle the demand on Election Day and our servers crashed. We hit one out of the park.
By 2012 we were able to set up a similar system for the presidential race and for every congressional race. Usage went up 50% and more importantly people spent twice as much time on VoteEasy and had twice as many page views compared to those using our general, far more powerful, deeper, richer Voter’s Self-Defense System website. Citizens simply did not want, or did not have the time to do the research. Because of our unquestioned credibility users just wanted the candidate’s position quickly exposed on the various issues of unique concern to them.
The experience was a revelation to Vote Smart. VoteEasy is just a tiny sampling of what we are capable of. We have and maintain all of the factual data on thousands of candidates and could go much deeper than just these few basic issues.
The lesson learned with VoteEasy was that the delivery of the Voter’s Self-Defense System is the key to citizen interest and use. It is not enough to be accurate, abundant and relevant, our delivery system must also be fast, easy and….entertaining! Plans for future elections will focus on developing new interactive tools that combine and synthesize the data.
Today, Vote Smart staff is busy developing coding systems by issues and topics for every piece of data entering the Voter’s Self-Defense System. They will then go back and code the millions of records already entered. This will enable Vote Smart to create far more sophisticated but entertaining, powerful but simple, arduous to create but instantly gratifying, data visualizations for millions of unique users.
What Vote Smart Needs To Do
Vote Smart is certain that VoteEasy has pointed the direction to the future of voter education but needs financial assistance in order to build a solid platform upon which future political data systems can be built.
In order to build that platform it is essential that Vote Smart do two things:
1. Code millions of bits of factual data by issue, by elected official, and by candidate (during election years). This must be done in two ways:
First, would be in Vote Smart’s work with the Texas Advanced Computer Center (TACC). With a $27 million grant from the National Science Foundation TACC possesses one of the world’s fastest Super Computers. Staff at TACC believe that it may be possible to create a program that is able to code like or associated bits of data within our enormous databases. These discussions will continue through March when we hope to have a working proposal.
Second, even if TACC is able to program the coding processes it will require extensive time from Vote Smart staff to review those codes for accuracy and usefulness in creating new data visualizations. In addition, we may find that TACC cannot successfully code political data. In that case, Vote Smart has created a plan that will do it manually using a team of research staff and interns at both its University of Texas and Montana research centers. Such a manual process would require a staff of 6 and 14 interns seven months to complete, but would still be done in time to create data-visualizations for the 2014 elections.
2. Creating data visualizations would have to take place after the coding is complete. The coding would open a whole new world for the nation’s voters where systems could be built to tailor information to each users interests. Here are some of the ideas currently under discussion:
Political Galaxy: An interactive application that would automatically combine like pieces of information in which a user has expressed interest. For example: If a user asks how Sen. McCain voted on off-shore oil drilling the tool would swirl together the galaxy of political data that Vote Smart has compiled lighting up those stars that represent information related to the user’s inquiry. Not only the answer requested but also other stars where related information exists would appear. In this case it might include contributions from oil or environmental interests, ratings from energy or environmental groups, public comments he made on off-shore oil, previous issue positions, or even biographical data showing experience in the subject matter. This would reduce the current need for up to an hour of research in our database to a few minutes.
Report Card: When a voter wants to see how a candidate is rated by various special interests they could go to our databases and instead of seeing long lists of digits representing more than 200 competing special interests, they would simply see a report card much like one all users are familiar with from their school days. They draw their cursor through the names of the special interests they are concerned with and the grades would instantly appear on the card, representing the evaluations done on them by those organizations.
Ocean of Cash: Again, instead of seeing endless rows of digits representing campaign contributions from interests, users would simply see a seething ocean of dollar bills. As they draw their cursors through the names or symbols of various special interests the ocean peaks in waves or drops into troughs, representing the relative amount of money received from each interest.
Candidate Cloud: A representation of each candidate’s preference for key words, phrases, or subjects of unique interest to the user, perhaps using clouds representing the user’s selections. For example, if “border fence” is a popular theme of the candidate’s it would show up as a large cloud. If “acid rain” is selected, by the user but is rarely or never mentioned it would be a tiny cloud or not appear at all.
Web/Mobile App: This involves the designing of all our interactive tools and our web site itself into systems that can easily be used and understood on hand-held devices. Mobile apps is an area Vote Smart has no experience in but realizes is essential to develop.