Palladio is a new tool for me, and I wasn’t sure what to expect. After loading the existing file into Palladio, I was presented with a graph that visualizes the connection data for the participants and their top ten selections of music from the Golden Record. For example, if two participants have similar music selections, they are connected by the specific “target” (music track), and regardless of where the nodes are, the relationship stays connected.

One of the first things I noticed was the location of my node and the connections. I observed that I was closer to the middle of the graph. Considering the tracks and nodes on the outskirts of the graph had fewer connections, this made me realize that I might have had more connections with multiple participants or selections of music.

The visualization itself does not inherently capture the explicit reasons behind why participants’ music choices are similar. It shows connections of data but does not explain the underlying factors of why the data was selected. While the visualization can help identify connections, it lacks the context needed to fully understand the selection of each participant’s choices. This visualization is truly only a snapshot of what participants chose during the time of data entry. The visualization may constantly change depending on who participates, personal interests, cultural contexts, etc.

In general, the data is helpful to identify selection trends, and understanding these patterns can lead to personalized music recommendations or marketing strategies tailored to specific groups, provided there is more information on the participants. For example, if we knew factors like age, gender, preferred genre of music, and experience with music, this may help provide more information on why certain tracks were selected more often than others. As I began playing with the data, the top 10 tracks that were selected:

1. Track 14: Melancholy Blues (16)
2. Track 7: Johnny B. Goode (16)
3. Track 3: Percussion (15)
4. Track 18: Fifth Symphony (14)
5. Track 5: Morning Star Devil Bird (13)
6. Track 6: El Cascabel (12)
7. Track 11: The Magic Flute (11)
8. Track 24: Flowing Streams
9. Track 9: Tsuru No Sugomori (10)
10. Track 25: Jaat Kahan Ho (10)

This is an image of the visualization when only the top 10 songs were selected, and nodes were modified depending on how many participants selected each track.

After selecting the top ten songs, I was curious to see how the data compared when seeing which songs I chose with a selection of participants. Out of the top 10, I only chose 6. From this, I went down the “label_1” facets to see which participants had the most similar options. Curious about how many similar songs we would have, I gathered the following:

• 4 songs that were selected: Chan Mi Lim, Duncan Hamilton, Jonathan Tromsness, Lucy Lai, Stephanie Takeda, Stephen Acree, Abdulehed Yarkin
• 5 songs that were selected: Julia Cheong, Katy Richards, Shannon Wong
• 6 songs that were selected: Warren Wong

However, even with similar songs and being selected from the top 10, each selection out of the top 10 varied between participants. I couldn’t determine specific factors of why we chose certain tracks or why we did not. It could be due to various factors like personal taste in music, criteria for selection, and cultural backgrounds. This shows that reflecting or interpreting the reasons for “null” choices in data visualizations can be challenging, as “null” choices often lack explicit information.

• ETEC 540