Page 1 of 1

Where should I go for gradschool?

Posted: Fri May 15, 2009 9:44 am
by Azrael84
Hi,

I'm a student from the UK about to embark on a PhD in the area of String theory/QG. I have a couple of options in front of me and I am struggling to decide which would be the best path.

1) University of Southern California (USC). The PhD would be much more "stringy" here, and I get the Impression (perhaps someone else could confirm?) that String theory in general is much more dominant in the US than Europe. The professor(s) I would potentially working with look very friendly and people I would enjoy working with. One issue I'm having is to try to ascertain how well respected this place is, all I can see is the rankings place it mid 40's for Physics gradschool, which doesn't sound too great?
Another thing about here is the PhD will be much longer than my other options (5-6 years), I will be glad of the graduate courses and support I will have to ease me into research however, as oppose to the UK PhD options which will likely throw me in at the deep in with a lot more of a "go read this book" attitude, vs formal lectures. I won't however be so thrilled at having to devote so much time away from research being a TA (20 hrs a week, vs almost nothing in the UK). Finally I really like the idea of living in California for a few years, everything I've seen of this place (I've not been there in person yet!) looks really great; Mountains, hiking, weather, climate, proximity to places I've always wanted to visit etc. So it could be a really good experience for me.

2) Nottingham (UK) QG group. It's really hard to compare how this institute stacks up against USC in terms of how well respected they are, and what my chances of getting a good postdoc after would be? The professors seem really good, and from publication lists it seems that they collaborate with Perimeter from time to time, which is of course v well respected in QG. PhD is only 3 years, and whilst that is good in many ways, I do wonder if without the formal graduate lectures teaching me QFT and advanced GR etc, will I be thrown into research somewhat out of my depth? Although no TA'ing which I view as a definite plus. Also Nottingham seems a lot less exciting than LA (although I don't want to sacrifice education for location, it could be a deciding factor if all else turned out to be equal).

3) Imperial MSc Quantum fields with funding. This is only an MSc whereas others would be directly PhD, but assuming I did well on this course and came within the top band, I could presumably be taken on as a PhD at Imperial (not sure about Oxbridge as Im guessing I'll still be beat out by part III, is this def the case?). Imperial is a lot more prestigious than both the above, and I would really like to take these courses as I feel I could really use some formal lectures in advanced QFT/GR/ Stringy things, and it would help very much when starting any research. It is a risk however since I'm not guranteed a PhD spot afterwards, or I may just end up with a PhD of the same calibre as I have now anyway.

So there is my dilemma, I would really appreciate any thoughts anyone has on the above, as I'm finding it impossible to decide. Thanks

Re: Where should I go for gradschool?

Posted: Fri May 15, 2009 5:02 pm
by Peter
I would go with Imperial. You will be in London, plenty of job opportunities around (hopefully by the time you graduate). If you find out that theoretical physics is not for you, then you still have a degree from a prestigious school. If things go well then you can reapply to better U.S. schools. I think there is little chance to find a theory job with a Ph.D. outside the top 30-40 (It is hard even in the top 10). If you were an experimentalist there would be more options though.

Re: Where should I go for gradschool?

Posted: Fri May 15, 2009 5:49 pm
by Azrael84
Peter wrote:I would go with Imperial. You will be in London, plenty of job opportunities around (hopefully by the time you graduate). If you find out that theretical physics is not for you, then you still have a degree from a prestigous school. If things go well then you can reapply to better U.S. schools. I think there is little chance to find a theory job with a Ph.D. outside the top 30-40 (It is hard even in the top 10). If you were an experimentalist there would be more options though.
Thanks, interesting response. The problem with Imperial though is that is only an MSc not a PhD there, I doubt I would get into Oxbridge on the back of it (vs Part III'ers), which means the only chance to up my current PhD offers is if I get taken on by Imperial afterwards. (I'd be pretty reluctant to go down the US route again after Imperial MSc since I'd be wasting my time during the first 2 years of grad courses, having done them all already). So it's a bit of a gamble that could end up just landing me in the same PhD position I have right now...although I really do like the idea of doing the courses on that MSc alot, they look great and London would be fun for a year.

So you would say USC is just too low for a theorist in terms of postdoc chances afterwards? This is something I was worried about from the rankings tables, is it pretty badly percieved for Physics in the US then? Don't suppose you know anything about Nottingham, UK?

Re: Where should I go for gradschool?

Posted: Fri May 15, 2009 6:58 pm
by Peter
Azrael84 wrote:
Peter wrote:I would go with Imperial. You will be in London, plenty of job opportunities around (hopefully by the time you graduate). If you find out that theretical physics is not for you, then you still have a degree from a prestigous school. If things go well then you can reapply to better U.S. schools. I think there is little chance to find a theory job with a Ph.D. outside the top 30-40 (It is hard even in the top 10). If you were an experimentalist there would be more options though.
Thanks, interesting response. The problem with Imperial though is that is only an MSc not a PhD there, I doubt I would get into Oxbridge on the back of it (vs Part III'ers), which means the only chance to up my current PhD offers is if I get taken on by Imperial afterwards. (I'd be pretty reluctant to go down the US route again after Imperial MSc since I'd be wasting my time during the first 2 years of grad courses, having done them all already). So it's a bit of a gamble that could end up just landing me in the same PhD position I have right now...although I really do like the idea of doing the courses on that MSc alot, they look great and London would be fun for a year.

So you would say USC is just too low for a theorist in terms of postdoc chances afterwards? This is something I was worried about from the rankings tables, is it pretty badly percieved for Physics in the US then? Don't suppose you know anything about Nottingham, UK?
(I am sorry for the mispellings in my previous post...)
It is going to be hard to find a postdoc in string theory no matter where you go. I think if you do not believe that you can get into a top 10 U.S. school (or a top 3 UK school) then it might be a good idea to give up on string theory and do something that is more down to Earth. It is very painful to realize after 5-6 years of hard work that you cannot find job. (Unless you are brilliant).

USC is a good school, but not particularly strong in physics. I have never heard about Nottingham, but this is my mistake.

Re: Where should I go for gradschool?

Posted: Fri May 15, 2009 8:15 pm
by Azrael84
Peter wrote:(I am sorry for the mispellings in my previous post...)
It is going to be hard to find a postdoc in string theory no matter where you go. I think if you do not believe that you can get into a top 10 U.S. school (or a top 3 UK school) then it might be a good idea to give up on string theory and do something that is more down to Earth. It is very painful to realize after 5-6 years of hard work that you cannot find job. (Unless you are brilliant).

USC is a good school, but not particularly strong in physics. I have never heard about Nottingham, but this is my mistake.
Yeah I agree it's going to be hard to find a postdoc in strings no matter what (the grad advisor I spoke to practically said as much too, and seemed to be swaying me over toward Quantum computing for some reason).

The reason I didnt get into top 3 UK school for a PhD, is that Oxford Maths Inst and Cambridge DAMTP both require (although not officially but certainly 99.99% of the time) part III and I can't afford to self fund this, and Imperial has a similiar prereq of their own MSc in Quant fields and fund forces (the one I have the option of doing now), which they typically skim PhDs from. As for not getting into a better US school, this was simply because my GRE was only in the 800's.....simply not high enough for an international aspriring string theorist. My academic record is pretty high, top 2 or 3 of year, publication etc etc, but perhaps this is nothing special in HEP-th.
However I do believe Nottingham is pretty good, despite Americans perhaps not hearing of it (I suspect most only know Oxbridge, Imperial, perhaps Durham?), one researcher in the QG is John Barret, who cocreated the Barret-Crane model, which seems like a pretty big thing in the world of QG.....at least as far as I can tell. Also the fact they seem to have collaborations with Perimeter makes me think it might be pretty well respected, since obviously that place is a giant in QG. I'm pretty sure if I went to Nottingham I'd be on track to get a post doc (the culture in Europe isn't quite as forocious as that in the US, plus QG not as insanely competitive as String theory).

I'm just a little bit sad about giving up the chance to experience a few years in California, and it would be nice to have some more formal grad courses instead of complete independant reading. I was hoping that USC would be at least academically equivalent to Nottingham, so I could have the best of both worlds.

Re: Where should I go for gradschool?

Posted: Mon May 18, 2009 8:53 pm
by grae313
Azrael84 wrote: (I'd be pretty reluctant to go down the US route again after Imperial MSc since I'd be wasting my time during the first 2 years of grad courses, having done them all already).
You wouldn't have to take any courses that you've already done with most schools. A student here who studied for a year abroad didn't take any classes and went directly into a particle theory research group. Nobody wants you to waste your time if you already know the stuff and are ready for research.

Re: Where should I go for gradschool?

Posted: Tue May 19, 2009 8:01 am
by kz205
I'd say go with the Imperial option.
I doubt I would get into Oxbridge on the back of it
These days, Imperial physics actually seems to be better than Oxford at least. In fact, one of my referees told me to forget Oxford and apply only to Imp and Cam in the UK. All of my theory friends here even prefer to stay at Imperial rather than go to Ox or Cam. Anyway, Imp, Ox and Cam are all good research buddies and they think very highly of each others students so there should be very little discrimination against you when you get around to applying for PhDs at these places, assuming you do decently at Imperial. I get the feeling that they still tend to favour their own students, but only slightly. The MSc is an added advantage since it is certainly much more prestigious than the standard MSci and BSc which most people, like myself, go for.