A little about myself. I am a masters student at cambridge university. I have been through the application process before, last year I applied for hep-th at eight top unis without offers. I am reapplying this year with more insight. I include this information only to provide context - I would prefer this discussion to be about my views rather than my personal position.
The first point I would like to make is that the aspects of the US application procedures I address are, to my knowledge, particular to the US.
Emphasis on research experience
Universities want applicants who will do well at PhD research. They want to differentiate between good and bad candidates. I personally believe that research experience is not a differentiator either in ability or potential to carry out research. What is the difference between an applicant A who has completed a summer project, and applicant B who has not? It is rare for someone to quit halfway through a project, so it is not a measure of an applicant's sticking ability. Nor is it I think a good measure of intelligence/physics ability, certainly not as good as the PGRE or undergraduate GPA.
So the difference is that applicant A got a summer project, while applicant B did not. Why did one get a project and the other not? I see that ability could play a part at this stage. Perhaps a supervisor chose applicant A over an applicant C because of superior GPA. But I think the role played is minor. The most important factors to getting a undergraduate research project are luck, who you know and applying to as many opportunities as possible. Are these a measure of an applicant's ability to carry out PhD research? I do not think so. There are other possible circumstantial reasons why one applicant got a project while the other did not:
a) Opportunities for research experience vary between countries. Science in the US is highly funded relative to other countries. There are more opportunities, such as REUs, for which there are no equivalent in most countries. It is not a fault of the applicant where they were born.
b) Priorities and culture. It is unusual for an undergraduate, at least in europe, to publish a paper. Even some postgraduates have not published by the time by they defend their thesis. The culture I feel in europe is that the PhD marks the beginning of your research career. Undergraduates are generally not yet skilled enough to carry out original research, most summer projects are menial and not reflective of a PhD. Undergraduates in the US are no more or less skilled, however they are more likely to pursue research experience, especially when they know the emphasis grad schools place on it. Applicants unaware of the different culture in the US are caught out.
c) Luck and circumstance. Two powerful driving forces.
These factors lead, I feel, to the conclusion that research experience is not a differentiator of applicants, and should play a small role if any in the selection process. The emphasis US grad schools place on it is not good. So how can adcomms differentiate between applicants who are identical in all ways except research experience?
Interviews
As an outsider, the fact that US grad schools do not usually interview their applicants is bizarre. It is like buying a house, without looking at it. I think US grad schools are missing out on much by not doing so. I know of two reasons commonly given as to why they do not:
a) The extra resources, especially time, required to interview candidates. I feel this is not a valid point. By offering an applicant a place, they are committing to ~6 years of teaching and training students. This is worth a 30 minute interview. If the number of applicants is too high, then a short list can be created from paper applications, and then interview those applicants. This is the method used by all universities, at least in the UK, not to mention all companies to select the best applicants.
b) How could the applicants be interviewed? Skype would be a good way. Many companies conduct interviews this way. If that is not possible, then by phone. What if the applicant is in a significantly different time zone? I know of no-one who would not stay up late for the opportunity of being interviewed by Harvard. Applicants will be happy, especially if they feel they are being given the chance to show more of themselves.
I think grad schools do not interview applicants because it is not the done thing.
GRE general paper
Most people I think would agree that the GRE general paper is not a good indicator of graduate physics performance. There have been studies which show this to be the case http://en.wikipedia.org/wiki/Graduate_R ... erformance So why do grad schools ask applicants to spend $195 and tens sometimes hundreds of hours on this exam? No reason. I personally find this particularly pointless and demeaning.
To summarise, if I were running a university's physics admissions procedure, I would create a shortlist of 2x the number of offers I plan to make, and then interview those applicants, 10 minutes general questions, 20 minutes technical questions. I would not ask applicants to take the general gre. I would probably put the aspects of a student's application in the following ranking of importance: Interview, grades, LoRs, PGRE, SoP, research experience.
These are all just my opinions. I'm interested in what other people have to say, for or against. Give it to me!
