1. When there is a choice between living in hotels and living with a current student, choose the latter.

Indeed, living in hotels or inns can be fun, espcially when it is for free. For example, I really enjoyed my staying at the lovely inn called Inn 1890 in San Francisco. Located near the Parnassus campus of UCSF and opened in 1890, it is a small inn with class and a history. Among the many things I like about Inn 1890, its kitchen on the 2nd floor is my favorite. Every morning, guests can enjoy free breakfast here, listening to classical music played from a stereo system. The egg pudding and Mozart's violin concerto really gave me a good start on the day of interview. I also wanted to see how's the new hotel at MIT is like. Murray mentioned it and told us that the walls of that hotel are decorated by printed circuit boards, a reflection of the taste of MIT. However, eventually I live at the dorm apartment of a current student in the program at MIT. At first I was a little disappointed, but after the process, I felt lucky because it really gave me much more chance to learn more about the program, and about the students' attitude toward it. Like my host Nick was, most student host of these interviews are very open, nice and helpful people. If they take an effort to provide you with a comfortable lodge, try to show you how satisfied and enthusiastic about the program, and care to give detailed answers to every question you raise, then you can be sure that the student is really having a good time studying here. However, if they don't (which is largely unlikely), you should consider the probability that there might be something wrong with the program, which you should find out by further inquiries. Another advantage of living with a student host is that you can learn more about housing opportunities on or around the campus. For my staying at MIT, I visited two different on-campus dorm buildings for grad students and was very impressed by them (considering that there are no grad dorms at Hopkins), especially the newly built Sydney-Pacific building, which was so elegantly decorated and well-equipped (There was a PS2 and a big flat-panel TV in each common room !!).

2. Prepare some figures or fliers for presenting your previous research projects.

At least for me, it is very difficult to talk about my research very clearly without visual aids. A few figures printed on A4 papers can greatly facilitate this process. In addition, it can impress the professors with how serious you are about their program and your own research. Don't bother to make those figures as complicated as conference posters, because probably you won't have enough time to go through all the details.

BTW, be sure to make good use of the limited amount of time during the one-on-one meeting with the professors. The professors sometime forget about the time limit (especially when they enjoy talking with you), and if that happens, you are the goal-keeper. Don't let the professor talk too much and give you too little time for asking him questions. You will be sorry saying bye-bye to him without finding everything you want to know about him.

3. Prepare not only for answers to specific questions about your research experience, but also answers to more random and general questions.

Indeed, random and general questions are much more difficult to answer than specific questions are. And they are also difficult to prepare for. It depends on your reservoir of previous knowledge, ideas and experience, and it also tests your ability to think quickly and put ideas into words on feet. However, you will do better if you are at least psychologically prepared for this kind of questions. For example, when I was interviewing at MIT, a student interviewer threw me a really general question: how does the nervous system process and transmit information. Of course, this was a question with no definite answers and it can be tackled on multiple levels. I didn't do well in answering it because I was not mentally prepared. (Fortunately the interviewer was not a faculty member :p) When I went to UCSF later, I did better on answering another random question like this, because I was aware of the possibility of this kind of questions springing out. This time it was a faculty member. (He was a student of Hubel and Wiesel, the nobel laurates, which I found out later. Luckily I didn't know that when I was interviewing with him :p). He asked me: what is the most interesting paper in systems neuroscience that you read recently. I thought for a while, and several possible answers occured to me. I chose the one that meet with his field of research best, and the conversation turned out nicely and he seemed to be satisfied.

4. Do a research on the faculty members you are going to interview with.

You should have gotten a list of the faculty members who are going to interview with you. If not, you should have in mind who will you visit. Don't go to them without knowing anything about their researches. This will put you in a passive position. Instead, go to their web pages and pull out the abstracts of their recent publications and print them out. You can read them on the airplane (BTW, it was my way of killing time during the flights). Then when you go to them, you can ask them more specific and informed questions about their researches. This can also impress them with how serious and well-prepared you are.

5. Spend some time to talk to the current students in the program.

If you have time, be sure to go to the dinners / parties held by the current students. As the student host, most students who go to these social activities are open and nice people who are willing to show you how enthusiastic they are about the program (and what they don't like about it). Most importantly, it is the one chance that you can pump the students for answers to the questions which are inappropriate to ask when faculty members are present, for example, whether certain professor have secure sources of fuding, whether he or she treats his students nicely, and whether some senior professor is going to retire soon, etc. Of course, sometimes it takes a little bit art of conversation to ask this kind of questions.

6. Go to the interview with a sense of expoloration.

There is no need to stress out because you fear that you might be rejected if you screw up the interviews. As the name suggests, interview is a two-way street. It also gives you an opportunity to learn how good the program and faculty are. In this sense, the professors should be more nervous than you are, because if they screw up, they will lose a student as good as you. Back to the track, be sure to learn as much as you can from the faculty members during the interviews, espcially those who you want to work with. These are the most important questions from my point of view:
1) What's his/her background, and what kind of career path did he go through. If he or she just view academics as a way of earning a living and not enthusiastic about research, then he/she might not be worth working with (in my opinion).
2) Is he willing to let the student do research in whatever they are interested in, even if it slightly deviates from his own line or research? If the professor thinks only about his own academic success and you are a student that likes to have your own creative idea, you probably won't feel happy working with him.
3) Does he travel a lot? Is he accessible on a daily basis? Does he spend a lot of time in the office and with his students? If he is always out of town giving talks / attending conferences, you won't have enough chances to discuss academic issues with him or develop a personal relationship with him, which is bad, in my opinion.

After all, interviews are totally fun. Go there with a light heart and enjoy it. And don't miss the opportunities to check out attractions in the city, espcially if you don't want to return there in near future.