Task: The instructor will provide a network database generated from the Golden Record Curation Quiz (Module 8) and post it below in a zip file. The database will be created by grouping your Quiz choices to display a social network derived from commonalities. You will be asked to visualize the database by loading the .json file contained in the zip folder as an existing project in the Palladio app. (Links to an external site.)
After loading the .json file in Palladio, you'll see something like this:
You are also able to select the different groupings listed in the "facets", which allow you to isolate the members or each group for further analysis.
For this task, you will analyze the visualizations and reflect on the implications and outcomes of their generation. The visualizations show not only connections between the participants' music choices, but also groups participants based on the strength of these choices, creating communities of individuals with similar responses. But exactly why are these responses similar? Is the visualization able to capture the reasons behind the choices?
Using these visualizations as prompts, reflect on the political implications of such groupings considering what data is missing, assumed, or misinterpreted. For example, while you may be able to justify your musical choices in the Quiz, there also may exist reasons why you did not choose other pieces. Can the reasons for these "null" choices ever be reflected/interpreted in the data? Post your analysis and reflection on your personal webspace and submit the URL. You can always produce supporting visuals by producing screenshots. Don't hesitate in contacting the instructors in case you have any issues in producing the materials for this task.
You can learn more about the Palladio project on the Stanford University website.
Analysis
These are the four groupings created by the app, which I will refer to as communities 1 through 4.
Community 4
This community includes my own choices, and, interestingly, all the students in this grouping are predominantly women, with the exception of Tom Watchorn. I thought this was fitting, since I chose my songs through a frame of feminism; looking for songs that were performed/written by women.
Another factor in this grouping is that the majority of the songs are orchestral, or include a strong multi-instrument component.
Community 3
This community only had three students in it, but the shape created was interesting. There is only one song that all three of the members share; Flowing Streams. But there are many many other paired overlaps. For example, Junel shares 5 songs in common with Yijun, who shares 5 with Helen, who shares 3 sith Junel. Each person also has a number of songs they do not share at all. 4 for Junel, 2 for Yijun and 3 for Helen.
All the songs Junel and Yijun share are wind instruments; trumpet and pipes.
All the music shared between two participants is instrumental.
I don’t have access to Yijun’s rationale, but both Helen and Junel chose their music based on “diversity”, so perhaps that is why these three have so much overlap.
Community 2
I found this community harder to analyze because it included more people, so the visual was more chaotic with five.
There were three common songs in this grouping; Johnny B. Goode, Melancholy Blues and The Fairie Round, but I couldn’t find much that linked these student’s choices together (musically speaking).
Community 1
The three songs this community shares don’t seem to have that much in common, other than that they are all very different. They come from very different cultures; Senegal, America and Azerbaijan.
Reflection
The instructions for this assignment say that the program groups students based on the “strength of their choices” (Peña, 2022). Certainly the algorithm cannot “see” the underlying music behind each choice, and thus cannot group them thematically in that way. Therefore the program is grouping us based only on the number of choices we share, it is an “unweighted network” with “unordered pairs of nodes” (Systems Innovation, 2015).
When I looked into the rationales behind the various members of my own community (4), we each had very different reasons for our choices, even though the tracks ended up being similar.
Kelcie chose my songs based on whether they were written or performed by women (Vouk, 2022), Jocelyn chose songs that had no reference to war or violence (Chan, 2022), Erin chose songs from around the world that showed “the progression of music over time” (Duchesne, 2022), Tom wanted to represent “all parts of the world, different cultures, time periods and types of music” (Watchorn, 2022), Kelly and her husband chose their personal favourites (Kelly, 2022), and Jenny carefully considered and chose each song individually (Schroeder, 2022).
This Palladio graph, although interesting, doesn’t capture the reasoning behind the choices, rendering it much less powerful than it could be, if not effectively useless. In my opinion, creating an unweighted network doesn’t really make sense for this kind of data. The really juicy, interesting part of this data set is why people chose what they did, not just which ones were most commonly chosen. It leaves the reader to fumble in the dark, trying to create connections out of thin air (as I did in my community musings above). Creating datasets in this way might encourage false conclusions, which in this case don’t particularly matter, but could have a much broader negative impact when applied to internet search engines, for example.
I wonder if the purpose of this assignment was to show us that a search algorithm that works only by showing the most common connections misses out on the fundamental importance of WHY two nodes are linked, not simply that they are.
References
Duchesne, E. (2022, July 7). Task 8: The Golden Record curation assignment. Erin Duchesne ETEC 540. https://blogs.ubc.ca/erinduchesneetec540/2022/07/07/task-8-the-golden-record-curation-assignment/
Kelly, K. (2022, July 10). Task 8: Golden Record curation. ETEC 540: Text Technologies: Katherine Kelly. https://blogs.ubc.ca/katherineetec540/2022/07/10/task-8-golden-record-curation/
Nelson, T. H. (1999). Xanalogical structure, needed now more than ever: Parallel documents, deep link to content, deep versioning, and deep re-use. ACM Computing Survey, 31(4). https://doi.org/10.1145/345966.346033
Peña, E. (2022). [9.1] What is the web and what is not. In ETEC 540: Text Technologies: The Changing Spaces of Reading and Writing. The University of British Columbia.
Schroeder, J. (2022, July 7). Task #8 Golden Record. Jenny Schroeder ETEC540. https://blogs.ubc.ca/jennyschroederetec540/2022/07/07/task-8-golden-record/
Systems Innovation. (2015, April 19). Network connections [Video]. YouTube. https://www.youtube.com/watch?v=2iViaEAytxw
Vouk, K. (2022, July 5). Task 8: Golden Record curation assignment. Education.Technology.Research: ETECetera… https://kelcievouk.wixsite.com/my-site/post/task-8-golden-record-curation-assignment
Watchorn, T. (n.d.). Task 8: Golden Record curation. Thomas Watchorn: ETEC 540 66A. https://blogs.ubc.ca/thomasw/task-8-golden-record-curation/
Commentaires