Wednesday, February 3, 2010

Thesis progress

My adviser says I'm done (almost). She recommended that I take what I've got and write a report on it, and that will satisfy the masters requirements. If this one last thing turns out to be significant, we'll spend next quarter preparing it to publish.

The last thing is to take a matched sample of non-bankrupt firms and compare their break point behavior to the bankrupt firms breakpoints for each indicator. If it's significant, we should see randomness in the non-bankrupt firms. However, as I'm typing this, I'm thinking that the time-frames might be an issue: the point of reference for bankrupt firms is the date of bankruptcy; the only thing I could use as a point of reference for non-bankrupt firms is the current date. Comparing the two might be confounding; I might get randomness as an artifact. We'll see.

I got a nice reaction from my adviser when I showed her the work I had done while she was away. She was having trouble interpreting a plot, I explained it, there was a thoughtful pause, then a big smile and: "This is really good work!"

No comments: