Grounded Theory

A reading list

There is a semi-official online definition at groundedtheory.com. That site is quite horrible. What is Grounded Theory? at Grounded Theory online is nicer. There's a 'How to do it' tab on that site which is a nice intro. The Resources page lists several books in the sidebar that seem to be ones we should acquire - though most seem to be available in the library at McMaster, so they should be checked out from there first! Having said all of that, I must say that the Wikipedia page is rather good, and is probably the best place to start.

From an SE perspective, it is worth reading Grounded Theory in Software Engineering (also in the private repo). Note Table 3 in that paper, which gives us a nice list of journals/conferences that might welcome our work. But also carefully note the 'Discussion' section, and the warnings there (of which we 'fall prey' if we merely use it for legitimizing our approach). While looking around, I found Grounded Theory for Geeks which I enjoyed.

I must admit that the part which is fuzziest to me is Open Coding. It basically seems to come down to finding commonalities, and then coming up with reasonable words to label them. See this example and this slightly longer discussion.

Discussion

What I am puzzled about is that I found no discussion of reconceptualization in the, admitedly cursory, look I did through the grounded theory literature. What if your search through the data was biased (for whatever reason), causing you to do a sub-optimal classification of the data and patterns? You might have not coded certain aspects of the data which turn out to be crucial. Or you might prematurely find patterns (codes for codes) that fit, but skew things in a bad direction. How do you recover from that? I think that a Kolmogorov Complexity influenced approach might really help here.