However, there is one big downfall of data-driven visualizations and that is the interpretation of that data. What is lacking is the STORY behind the data. In this case, what is the reason WHY people made their choices? Visualizations can absolutely show relations between data that we would never see on our own, especially in this era of Big Data but it is the interpretation of the meaning behind the data that is less transparent. For example, in this exercise, where we are visualizing the curation of the 26 songs on the Gold Record down to 10, why did I choose such a high percentage of the same songs as the people in my group, more so than in the other groups? What cannot be derived from the numbers alone is the reasoning behind why we chose the same musical selections. Did these four classmates have similar reasoning in that they too were curating tracks through a multi-cultural lens with strong emotionality or was it something completely different? Were they just songs they liked? Were they musical selections that tied to something in their past or a million other reasons?
Variables I wish that our dataset included demographical information so that we could run some hypothesis through like ‘Are women more likely to choose x song’ or ‘Are people more likely to pick songs from countries they have travelled to?’ I have listed out some possible reasoning I considered but don’t truly have a way of knowing any relevance without doing further data collection.
Another thing I grew increasingly interested in was the fact that one song ‘Men’s House Song’ only had one connection (only one person chose this track). Why was that? Why did so many of us decide to overlook this song but it was deemed culturally significant enough to be included in the original 26 song list. The final piece that was unclear to me from the visualization was whether there were any of the song selections that had no votes or a null response. I’m not sure if it would be left out of the visualization entirely or it would be an orphaned data node floating to the side. I believe I counted all 26 songs represented but I didn’t see a way to confirm this easily without manually counting them. Reasoning After brainstorming this list by looking at the data, I decided to snoop around the web spaces for people in my facet (grouping) to see if any of my possible reasonings could be confirmed by reading their Task 8 posting. In doing so I found that there were serval selection criteria that we had in common.
There was a lot of nuance in how these criteria were used or the weight we put on each but this may have lead to our similarity of lists.
2 Comments
8/11/2020 11:11:17 am
I think it's really interesting how you discussed the possible variables, in particular, the one that relates to gender. In my network task I focused specifically on gender, grouping males and females and their selections separately to determine if there was any underlying connections based on that information. While I found some reasoning, I too, needed more information on the story behind the choices and could not gather this data from Palladio. It would be interesting if our reasoning for our choices was also added to this and what degrees of connectivity we could then see.
Reply
Tanya Weder
8/12/2020 09:22:47 pm
Agreed Sasha. I really liked the mixed method approach where researchers get both the power of numbers and the story behind it.
Reply
Leave a Reply. |