If your time frame is a 1 year MSc, I would be very wary with setting up a project that requires both a very hard to find sample and tricky ethics to clear. Unless you already have supervisors or other stakeholders with access to (ex)offender populations, and already have their buy in, it is highly unlikely you are going to be able to recruit enough numbers.
Although I now tend to supervise doctoral projects with a much larger timeframe, I have done the odd clinical MSc project, where I already have a patient group, NHS clinicians and an existing data methodology for service evaluations etc. Even with those things in hand MSc students have suffered as they have to develop their proposal and get their head around the methodology, then clear ethics and the beaurocratic hurdles. For MSc, simple is always better. Piggy back of an existing larger research project, do something that is achievable in a limited timeframe and use a tried and tested method. If I had MSc students again I would now recommend replication studies rather than doing a completely novel approach.