What about the projects themselves - which one has the better long-term potential? Very tough choice.

The reverse is also true in some places. Where I am, it is often taken for granted that grad students and junior academics will just soldier on and do whatever it takes, no matter what. There are a few casualties but the general culture is that if you're any good, you don't need supervision. It's not always like this - but it happens a lot - too much.

It's not the MAJORITY of the grad students I've ever known, frankly.

I bet my data is uglier than your data....

I keep tweaking my theoretical models. I can't seem to help myself it's likea disease. I keep thinking their finished and then I start tweaking again. And no matter how small the tweak is, it means doing all or a lot of the analyses over again. It's driving me bonkers. i'm about to do them for the last time. Yes. The last. Really. No more tweaking. None. Not even a tiny tweakette.

I think this is common then a team of people work on an overall project. I don't, but there are a couple of big projects like that in my dept. Clearly some people are more in the loop than others. The important thing is not to cede your project to the others, inch by inch. Now that you see the danger, you can better forstall this in future. Be clear about having your own bouned chunk of work.

This may be a rubbish suggestion, but is it possible to contact you first supervisor and ask them to give you feedback on your work?

I think it depends on which differences you are talking about. Some differences are easy to weather - but some cannot be born. If the conflict is such that the two views are totally mutually exclusive in a topic which is central to your supervisor's work, that could be too difficult and begs the question why you are working together at all.

I published a paper without my supervisor for this reason. There was non acrimony at all, I just realised and pointed out that we had stalled due to a conflict over some of the inferences, he agreed, and I published on my own. He was just happy to get another paper and not at all bothered.

I'm about to have the same problem again - but this time is much more serious because some of my data is 'owned' by another researcher and I actually cannot publish without their approval. This is why I haven't done it yet - I'm still figuring out the best way to support our differing views in the one publication. I think it can be done - we'll see.

Some supervisors help/interfere more than others. Sounds like yours is a bit too 'helpful'. My supervisor will occaisionally offer a casual opinion or suggestion - but never tell me to do X or Y. It seems there is rarely a happy medium.

It would be just the same with the Mres - you'd still need a course in stats. When you start your PhD there should be some relevant course(s) you can attend even if they are in another dept.

Yes, but you may find that they are reluctant to take the exact same project. You would have to find a very good match. But in theory, it should be possible.

It can stand alone bit there are two reasons I would want to continue at the same institution: a) if you continue with the same research project you can possibly do the PhD in 2 rather than 3 years and b) it is common practice to stay on with a supervisor from your MPhil who you already know, warts and all, and who already likes you project area. it is a very standard way of ingratiating yourself with a suitable supervisor who comes to know and love you rather than emailing and having a brie meeting with a new one.

I expect my lit review to be the introduction - but the methodology to be a different chapter. Unless your thesis is a set of seperate pieces with their own lit reviews, in which case you would have a very short overall intro.

An MPhil can be part taught but should have a research component. The research component can take the form of a dissertation that reanalyses/meta-analyses other material. In any case, it is not equivalent to a doctorate. Sometimes the research component can be expanded and carried on into a PhD.