Comparing across sessions (part 2)

Following up on this post.

I was able to spend some time this weekend working on this, and got about a fifth of the way through the first round (shaping analysis). What was enjoyable was working with the material in Compendium. I want to be able to do all sorts of comparisons between the sessions as well as look across and through the data in unforeseen ways, which doing this part of the analysis in Compendium will help with, as well as potentially to be an interesting, web-accessible way of giving others access to the materials.

First I went through all of the 'shaping forms' (one for each session, as you can see here), and looked for aspects that appeared to recur throughout each of the session (i.e., dimensions, according to grounded theory). I put one dimension each on index card, then sorted them into groups.


I then typed up all of the dimensions in a Word doc, refining the descriptions as I went, then put them in a rough order within each of the five groups (the doc is temporarily here).

Next, I imported all of that into Compendium. I then worked through the six dimensions in the first of the five groups (Group A: "Aspects having to do with initial plan and other pre-session aspects, such as choice of method and approach").

Doing this in Compendium helped me to further refine, order, and sub-group the different dimensions. I was refining the approach as I went, so not all the 6 dimensions I got through are fully consistent yet, but I came to an approach where I would do the following for each session within each dimension:

• characterize where each session lay along the high-to-low (or other values) for that dimension
• give a justification/rationale for why I gave a session that value, captured as a Pro
• array all 8 of the sessions along the 'axis' for that dimension, generally with the 'high' at the top of the map and 'low' at the bottom

You can see the interim results at these links: With left nav menu or Full page view. This map is the most interesting one, at least visually, since it has clickable thumbnails of the artifacts it refers to.

This is the kind of analysis I envisioned doing when I started this work six years ago. More to come.