As a first step I cleaned the data and reduced some of the dimensions. I found the information about what one is hoping to get out of the class irrelevant to group formation, therefore I completely removed it. Initially there are 9 dimensions of the quantitative data. This can be reduced to three main dimensions: mathematical/data skills (IVIS, Stats, Math), computer/programming skills (User, Prog, Graphics), artistic/user facing skills (HCI, UX, Art).
The x-axis represents the total average score for each student while the y-axis their score in their strongest domain. Colours indicate which domain students belong to in terms of their strength.
As data is based on self-reporting I thought exposing uncertainty in the data is important . For students who have extremely high or low overall average this might be an effect of difference is self-confidence and therefore their highest score has to be looked at with an eye on their overall average.
This representation allows the performance of most low-level components of analytic activity (retrieve value, filter, find extremum, sort, determine range, distribution) as defined by Amar et al. .
The solution follows the Visual Information Seeking Mantra: overview first, zoom and filter, then details-on demand (zoom is unfortunately missing - lack of time) . One can filter which domain they would like to see on the visualisation by clicking on the respective button - this way allowing better comparison within the domain.
Details on demand is implemented in a tooltip style: when hovering over a node one can find extra information: this case the Alias and the person’s interest. To allow better access to the details the node also changes color as we hover over it.
Shneiderman, Ben. "The eyes have it: A task by data type taxonomy for information visualizations." Visual Languages, 1996. Proceedings., IEEE Symposium on. IEEE, 1996.
 Amar, Robert, and John Stasko. "BEST PAPER: A knowledge task-based framework for design and evaluation of information visualizations." Information Visualization, 2004. INFOVIS 2004. IEEE Symposium on. IEEE, 2004.
 Amar, Robert, James Eagan, and John Stasko. "Low-level components of analytic activity in information visualization." Information Visualization, 2005. INFOVIS 2005. IEEE Symposium on. IEEE, 2005.