This week I had my first meeting with my Futures Project supervisor, Stuart King. This raised new questions (but of the interesting and exciting type); things to think about from an initial idea that so far, seems “fertile”. To recap, my current idea for my Futures Project is a data visualization that shows the educational trajectories of specific people in Uruguay and tells their stories, mainly focusing on those who dropped out, what their lives are like, why they dropped out, and what their experience in the education system was like.
In this blog post, I will list some of the questions and concerns and then delve further into two specific aspects that require deeper examination.
- What am I going to plot? If I made a line chart, as is my initial idea, I know the independent variable would be time (measured in some way or another). But what would the dependent variable be exactly? This also raises the concern of how to show “bad examples” or trajectories that didn’t go “ideally”. These would be people that I will have interviewed and will have agreed to partake in this project. So I would have to think ways in which to portray them that preserve their dignity and don’t show them in a bad light. Even thinking about the Dollar Street project, which serves as an inspiration for my project, I don’t agree with many of the portrayals and communicational decisions made for this platform. For this, I thought of seeking advice from my former supervisor at UNICEF, who is superb at communicating development projects and stories in an empowering way.
- Would a line chart be the best form of representation? As mentioned, maybe it’s the linear way of representation that’s problematic, and I could explore other methods (e.g. tree diagrams) to depict these trajectories.
- Would I include a speculative/future-thinking element? While the data visualization would depict real stories of the past and present of these people, would I want to add a prospective element to picture how the future of the stories could potentially look like…? How would I build this (maybe including the protagonists’ input) …?
- What message do I want to convey? Apart from telling these individual stories, what would be the take-home message that I would want the audience to leave with?
Data collection
This was one of the aspects we talked about the most during our meeting, and so I thought it deserved a section of its own.
Considering that this is a very bespoke project, it needs a data collection of its own. I think using a preexisting database wouldn’t suffice to be able to construct these stories. This brings about ethical concerns for which I would have to submit the ethics form for approval at university.
Some of the questions raised during our discussion were:
- What organisations/people will I reach out to to collect the data? I have some connections through my work at UNICEF and volunteering for a charity that works in education.
- Would I do a survey (easier to get responses) or an interview (would provide more thorough data)? I would conduct short interviews (15-20 minutes) to gather the insights I need since I believe a survey wouldn’t provide that much information.
- How will I structure the data collection (e.g. What would the interview schedule look like?)
- How am I going to collect the data from the UK? Maybe conduct Zoom interviews? But still, I would need someone in Uruguay to get in touch with the interviewees.
- Would I interview adults or adolescents? I think interviewing adults would give me a more long-term perspective on their educational trajectory. But I think finding adults would be harder since they usually lose touch with charities I could potentially reach out to once they become adults.
- How many stories will I aim to collect? Somewhere around twenty…? However, this would depend on the quality of the interviews and the data collected. Also, how would I ensure there’s a certain representativity, especially regarding geographic locations within the country?
- Ethical considerations especially make me question the level of personalisation of the stories. For example, would I use a photo and a name to give these stories in the plot a human face (which is partly my aim with my project: to imbue humanity in the data)? Would I not use a photo and use a fake name? (This would affect the nature of the project) Could I do something in between, like using an AI-generated photo or the silhouette of the person?
Audience
I envisage the audience of this data visualisation to be the general public (i.e. as if they read this in a newspaper) and decision-makers. My supervisor’s position was that decision-makers tend to prefer broader statistical studies that provide the big picture. While I do agree, I also think that the country already has this; what it’s missing is getting to know these individual stories. Still, this statistical, broader information could be included in the data visualisation for context.
Another issue raised was that of language. The data will be collected in Spanish, and what I had initially thought was making the data visualisation in English for the purposes of the Futures Project and afterwards building a version in Spanish for local dissemination of my project and portfolio building. What I hadn’t fathomed was that translating the stories from the data in Spanish to the finished product in English would add another layer of interpretation and potential distortion of the interviewees’ accounts. On the other hand, it is true that the target audience would consume this piece in Spanish, so making it in English would be solely for the purposes of the Futures Project. For this reason, Dr. King suggested I make it in Spanish, but I remain sceptical about it. Still, we would have to formally check with EFI to see if this is possible.
Technical aspects
I envision this data visualisation to be web based, where the user would have an overview of all trajectories intertwined, and then the possibility to see “details-on-demand” (Shneiderman), by hovering over the stories and clicking to see more details.
I do not currently have all the skills to put this idea into practice, but I believe that with the practice from my Representing Data group project and with the help of Stuart King for my Futures Project, I can make it possible.
Some of the tools we discussed for this kind of data visualisation were D3, and Python (Altair, Vegalight), but also the practice of sketching and finding something that I like to reverse engineer.
Next steps
The meeting was very productive and left me with a lot of questions and thoughts to move forward. I have already set up a meeting with an education expert who I worked with at UNICEF to discuss organisations and ways to tackle data collection.
I will also look at the ethics form and inspiration from other projects to move forward.
Leave a Reply