Finding something to focus on for the research is proving hard work. There’s lots of things I could investigate just because they’re interesting or new for me but for the PhD that’s not criteria enough. I’m trying to find something that:

• I already know lots about or feel I could quickly grasp
• is interesting
• will make a new contribution to knowledge, just writing a textbook isn’t on
• can actually be answered, so no “meaning of life” type questions and isn’t likely to be answered in the next few months
• passes the ‘so what?’ test
  • Formally there are quite a few such requirements to meet and it’s proving tricky to meet them. So far I’ve asked some questions, searched in the literature about them and found them already answered or at least having some recent work. In a way that’s good news ‘cos it means I’m asking good questions, questions that others also ask and have only recently tackled. I was hoping to see if using network-based metrics improved fault prediction. It seems the answer is “yes they do, but not much”. And fault prediction looks increasingly like a done topic with each new development adding only tiny improvements. Plenty to write about for a new textbook, but see the above list… At the suggestion of my supervisor I’ve been looking around in other topics I know about (such as Physics) to see if I can create some fresh ideas. Actually that’s not true - I don’t lack ideas! But the need is to be realistic and extend existing ideas not wild out-the-blue stuff, so casting widely is more about seeing if I can apply concepts from one discipline into another. So far it’s proving helpful, some proposals are getting written and due to be discussed with supervisors before Christmas.