• offering some purely empirical number-based study on whether training courses for adults do or do not reduce the amount of cycling deaths on the road

    If you could supply that with a reasonable degree of certainty within the restrictions of an MSc thesis, a lot of people would want to talk to you :)

    As I said, your approach depends a great deal on what you want from it. Interested in more accedemia, then sure, you can go more theoretical. Interested in a specific industry or profession, give yourself a tight, original brief you can answer effectively.

  • Yes - in the case above it would be showing your ability to gather and rigurously process the data in such as way so as to produce defendable and conclusive statistics. I'm not talking a meta-analysis, although if you just want a pass that could be a way to go.