Week 7 summary: data visualization

For this summary post, I thought I’d reflect a bit on data visualizations so that I could record how my thoughts might change after the coming week’s discussions on algorithmic cultures.

I will try to respond specifically to the thoughtful questions Jeremy posted about my ethnographic artefact:

...my question would be about what you think was lost and gained through representing the community in this way. A more traditional ‘ethnography’ might have generated written field notes, so do the visualisations add something more ‘objective’ or ‘evidence based’ here? And perhaps more generally, how do you view data visualisations – do they provide the new ‘truths’ about social life so often promised?

The data visualizations provided an elegant solution to my own concerns about representing private conversations publicly. This was my foremost reason for using them. The second reason was the format is visually engaging and therefore offers a good starting point for discussion.

But as I’ve seen in the comments to my artefact, the images cannot stand alone by themselves. They require explanations about how they were generated (as Jeremy said, the format needs to be interrogated) and clarifications about what they exactly mean. In my introduction to the artefact, I acknowledged that while data visualizations are seemingly objective, they are actually quite subjective. The subjectivity I referred to is about what I chose to visualize–a particular discussion forum within the MOOC–and about how I highlighted specific features of that forum. After reading the comments to the artefact, I realized that the data visualization introduces another form of subjectivity through the algorithms it uses. For example, the tool I used prefers longer text, requiring 5,000 characters as a minimum; otherwise it does not work. Furthermore, the tool, in my view, seems to prefer text that is not just longer but also coherent. The example visualizations used in the site are Wikipedia articles. In contrast, the text I entered were relatively short discussion posts, each one made up of three to five short paragraphs on average. I wonder how the disjointedness of the text I entered affected the visualizations the tool came up with.

This brings me to my belief that data visualizations may help uncover insights about data, but those insights need to be verified always. Data visualization is only one type of tool.

 

The rest of the week, I posted a few comments to my classmates’ artefacts. The related material I posted and tweeted for the week focused on the different types of MOOCs, the c and x varieties, plus the more obscure ones whose acronyms made up a veritable alphabet soup.

On the two types of MOOCs

Although set apart by a few years, these two videos to me seem typical of how the main proponents of xMOOCs and cMOOCs frame their accounts of the history of Massive Open Online Courses.

It is quite telling how the two types do not seem to acknowledge each other. On one hand, Dave Cormier in his account of MOOCs does not mention that the popularity of MOOCs is often attributed to platforms like Coursera, EdX and Udacity, and not the cMOOC variety. On the other hand, Daphne Koller does not acknowledge the pedagogical innovations, both early and continuing, conducted by cMOOCs. While xMOOCs have gained a mass audience, cMOOCs are often attended by educators.

Why is it that these networks have never crossed or combined? I wonder if it would be possible to combine the reach of cMOOCs with the innovation of xMOOCs. If anyone knows of a video with both cMOOCs and xMOOCs proponents has been posted online, please let me know!