I agree with those who say move on. If you stayed (and that is if they were willing to have you back) there is a likelihood this will affect your time in the department, particularly if you complained about the head of department who is still there. How do you know that further issues won't arise in the future?

It's unfortunate, and you may feel it unfair, but I that's how it goes sometimes and I do think you'd be best off starting afresh somewhere new.

Also, just to clarify, was he told that the universities would take him by the general admissions offices or by specific departments/supervisors?

If the former, then they are talking generally and would not be in the position to offer funding - that should be discussed with a specific supervisor/department.

If the latter then it might be worth asking whether funding may become available in the future or seeing if any advertised PhD positions come up.

Depends on the field. In science, PhD offers usually come with funding. In arts and humanities this is less common. This is irrespective of one's previous qualifications.

Not being offered funding by the university does not mean there is no possibility of funding - there are research councils and charitable foundations to which one may apply. One may also proceed unfunded, but I would strongly caution against that, at least not full time.

What are your friend's reasons for wanting to do a PhD?

I agree with CR1980, it's really important to have interests and things in your diary outside of the PhD.Either a hobby you've let slide or taking up something new - but preferably one which will get you interacting with other people. Search online for local groups doing different things e.g. http://www.meetup.com/

It's also important to get peer support for your PhD, as it is always a lonely process, even when you have a desk in an office with others. Ideally this would come from your uni, but as that may not be practical, look online. This forum is a good source of general support. In terms of your own topic, set up a Twitter account for academic purposes, follow relevant academics, PhD students and departments, and after a while you will develop your own, highly relevant 'virtual' network.

Assuming you're right and that it is somehow possible to gain a credible PhD from a credible university in a shorter period, and that all the PhD students who take 3+ years are somehow slackers, less efficient, dawdling, not trying hard enough, don't have your insight etc.....

1. If you want to do a short PhD, you have to do one in something that you already have expertise in. To think that you can just walk into another field and start quality research in it without spending time laying down the foundations is either arrogant, naive, or both.

2. If you want to do biosciences despite my advice above, do not go anywhere near animal models or even cell culture. It's probably one of the most laborious and time consuming ways of gathering biological data. Give yourself a crash course in bioinformatics, find a lab with a tonne of data sitting around, and get analysing.

It's also worth noting that most good universities wouldn't actually let you submit before a certain time period - e.g. not before you've done at least two years of actual research.

One other thing - I would also advice getting some external support whether from a student counsellor or a trusted friend. It sounds like you are very distressed and could do with some emotional support as well as the practical advice.

On paper I'd lean towards candidate B. However, the stepping on toes point is worth considering, particularly is candidate A is head of PhD studies.

Is there a way to get them both involved? For example, at my uni, we are supposed to (ha!) have two supervisors, plus an advisor. The advisor is supposed to give support more on the process than the particular topic. If that's the setup you have, I'd go for B as supervisor, A as advisor. But it does depend on internal politics.

If you really want B, perhaps you could explain your dilemma with him (being careful!) and see if he can advise on how to handle it.

I agree that for small sample sizes, percentages might not be helpful. But I was trying to make the point that 'some' sufficiently ambiguous as to not be helpful either. If you have 20 people and 'some' people express and opinion, to my mind it makes a difference if that 'some' represents 2 people or represents 18. It may have limited generalisability, but it's worth knowing if something is expressed by the majority of your sample, or only a couple of people.

I do agree that the underlying themes are important, particularly if those opinions expressed overlap to some degree

That is a very specific topic. Could you widen your interests out a little?

If you have good programming/analysis skills I'd suggest getting into bioinformatics/health informatics. There's a lot of data accumulating out there that isn't being fully exploited.