Design for Understanding

The design objective of this assignment was to create graphics to convey data in two different perspectives. One was to give the unbiased view of the data; the other was intended to persuade the audience of a viewpoint. The data set we worked with was related to the famous show ‘The Bachelorette.’ We anonymously voted and eventually decided upon ‘The Bachelorette’ due to the large dataset, as well as the general uniqueness of the topic. By the end we had five different graphs.

We split into two teams- one to make the persuasive perspective and another to present the data bias free (I was part of the latter team). For the bias-free perspective, we decided to analyze the data from different perspectives- such as the effects of the number of roses, the first-impression rose, and the number of dates. The biased perspective intended to convince the reader that there was a gender difference between the Bachelorette and the Bachelor. The pros of the unbiased graphics is that there isn’t an agenda behind it, so you can trust the graphs more. However, the persuasive graphs have arguably more interesting results.

For the first round, Xiaoyue and I made a total of three graphs. We struggled a lot because we had trouble with the software. P5js wouldn’t work at all for me, so we tried Vega-Lite. While we knew what we wanted, trying to get Vega-Lite to produce that proved difficult. Nonetheless, we made a few graphs. We got a lot of great feedback. They really liked the information the graph conveyed (found it interesting or contrary to what they initially believed) and liked the interactive parts (such as the fact that hovering over gave further information). However, there was a lot of room for improvement.

Many people commented that our initial graphs didn’t have titles (and some even lacked axes). They also noted when the phrasing of the axes was confusing and took away from the message. Finally, they highlighted the fact that the graphs were far too small. With this feedback in hand, we went back and further edited the graphs.

The second time around, we implemented a lot of changes- we added titles, rephrased axes, and made the graphs bigger. However, while we had really good feedback, we didn’t really do a very good job of implementing solutions. One key point was the axis labeling. Many had pointed out they were either intuitive or confusing. While we did make changes to the axes, I don’t think we ever really got to the heart of the problem. As a result, we had different axis titles, but they were equally as confusing.

The second user testing yielded better results, with people understanding some of the graphs better. However, one main comment still stuck out- people were having trouble understanding our graphs easily. This was due to two things- confusing portrayal of the information and lack of background knowledge of what the Bachelorette entails.

After this round, we went back and made some more adjustments. Our final graphs ended up looking something like this:

Unfortunately, this project did not go as successfully as I wanted it to. I think the two things that stunted the team was our lack of familiarity with the software and lack of coordination in the team. None of us had any idea how to use the software, and we spent quite a bit of time trying to teach ourselves (a few other groups I knew had at least one person who knew one of the softwares well). Additionally, for some reason, we didn’t work as effectively as a team we could. I also feel I could have done a better job as I could spot where and why the graphs were confusing, but didn’t really speak up as I didn’t know how to make them “not confusing.”

Given more time I would have focused more on the team working together to utilize each of our strengths. Perhaps we could have gone to TA hours to get help with the software. We could have also perhaps used a different data set (one that needed less background knowledge). I think we also could have used the 5 page method more closely (we only loosely followed it).

This was very disappointing for me as Data Analytics and how we can more effectively convey data is a huge interest of mine. I wish the outcome could have been better. Nonetheless, I learnt a lot and I might practice more in my spare time.

Related posts:

Over the course of this project, partnered with Matthew Puentes, Ethan Chang and Jason King, I worked primarily on the “persuasive” component of data representation. During the first class our group…

As wedding officiants, we’ve got a lot of options when it comes to getting our wedding ceremony script down in black and white. I recommend Google Docs. I’m not going to go into all the particular…

On last tuesday I met with my friend Ahmad Raza at his place to share somethings which I knows but he didn’t know that before. When I briefly told him these steps he become inspire and asks me where…