Coming from a final year PhD student (already passed "pre-defence") doing a three year PhD in Japan:
Advice I would have liked: Don't do a PhD if you like taking classes and just want to extent that experience, as doing a PhD is very different from taking classes. Also, the classes you are forced to take probably suck.
Advice I received but ignored but is probably true: Doing a PhD actually decreases your chance of getting a job (especially in Computer Science)
Advice that I would have liked to receive, know now but still ignore: If you are in a foreign country and you don't know the language but you want to stay in the country after you graduate, learn the language! Don't spend all your time doing research thinking that you will somehow compensate for not fluently speaking the language just because you have more research achievements. (Disclaimer: I live in Japan, and most jobs, even academic jobs, require a sufficient level of Japanese language abilities)
Is it actually feasible to learn Japanese to a professional proficiency level as a foreigner? I've studied Japanese for several years, and can speak Japanese at an intermediate conversational level, but I'm still a long way off from communicating at a native level. So it seems like a tall order, especially given that you're busy with research and can't dedicate time to learning the language.
Starting from scratch is indeed a tall order, and probably not possible if you didn't study it before coming here. Many students spend some time as a "research student" before starting with the PhD, so that would be the best time to get a head start.
Its mostly what I hear from the postdocs in my lab. In Japan older companies seem to prefer to hire a more homogeneous work force.
Also in The Netherlands some tech companies (including one I interned at) were reprimanded for having a hiring process that discriminated against older applicants. The company I worked for had as requirement in their job ad "graduated less than 3 years ago".
Advice I would have liked: no matter how successful you are in grad school, its still going to be extremely difficult/maybe not worth it to get a tenure-track faculty position and there is a large opportunity cost for the 6-10 years spent in grad school + postdocs.
I would add that you should choose a lab with lots of funding.
Big labs might mean you don't spend much time with your on-paper supervisors, but you will get access to post-docs who can mentor you better than any professor who doesn't step into a lab. Big labs with funding also have lots of machines which can help you churn out data. Churning out data is very very important!
Source: I failed my PhD partly because there was no one to mentor me, and no resources to produce results. As a PhD student, you don't know what you are doing and you need those resources so you don't waste your time.
oh noes, not that guy again! this was written over 4 years ago before i became a professor, so YMMV :) i haven't looked at this for years; this is a more updated (and complementary) take on some of these issues, how that I've advised ~25 research students so far:
http://www.pgbovine.net/managing-me-as-your-advisor.htm
1. Keeping a good notebook means not having to repeat work because you can't remember what you did. Being unable to reproduce my own work was a massive handicap.
2. Your results "decay" over time. Even if nobody is counting your hours, the level of intensity required to overcome that rate of decay is intrinsic to the nature of the research itself.
A potentially useful clarification on what good means: on the one hand it should be easy for you to write to it, on the other hand it should be possible for you to search through it. I find myself re-deriving the same gradient over and over...
I think we are over hyping PhD progams. Especially in computer science.
From what I saw after graduation: The top 10% of the students got huge offers to join big companies.
The "Street smart" students are usually joining startups.
The most interesting part is that students joining a PhD program were not the best academically, but they were the ones that loved to brag about how difficult their PhD were going to be.
This is very US-centric, your mileage may vary in other countries. For example:
>For instance, I started my first successful Ph.D. project at the beginning of my fourth year and didn't get the paper published until the middle of my fifth year.
If you're an international PhD student in Australia your student visa ends after 3 1/2 years.
What I would add:
- Randomness
You're new, so you have no idea whether your project will work out. Most don't, that's just how research works. Some supervisors have a magic nose when it comes to evaluating a project, but even they occasionally fail. One of the best things you can learn in your PhD is when to pull the plug on a project that's just not working out. For me it was very frustrating to see other PhD students who 'stumbled' into having more papers than me, some projects immediately throw up results once you go near them and you get heaps of papers, some projects have to have their results pulled out of them by force.
- be pro-active
Most supervisors don't regularly keep tabs on their students, they're busy themselves. If something is unclear immediately find your supervisor and ask. It's better to over-ask then to under-ask and do the wrong thing for a month. I tried to talk to my supervisor at least 3-5 times per week (when he was in his office).
- write, write, write
In the current funding system the only thing that counts is publications. Start writing as early as possible, ideally have an outline of the paper when you're starting your project. Keep on writing in the paper even though you don't have the results yet - you want to 'tell a story', writing the story will tell you which experiments or analyses you have to run, and which graphs and tables will support your story. If your story doesn't work out change the story later. Writing the paper will also keep you from doing unnecessary work that sounds interesting in itself but won't add to the story.
- hard workers are more successful than geniuses
Guo touches on this a bit. Someone who works hard will have inevitably more success than a lazy genius.
- start communities if there are none
For software people, many unis have Hacky Hours now - often there's a local Software Carpentry community or similar. Get involved with those people, at the least just to get out of your office/lab once in a while.
- Alternative careers are myriad
Chances are you won't become a Professor. Depending on your field there are many, many hidden opportunities outside of academia, so look around!
Addendum to 'be pro-active':
You'll quickly realise that your supervisor doesn't know what's going on in the entire field, nobody does. Be pro-active in seeking out new developments and new papers that touch on your field. Set up Google Scholar alerts, set up PubMed alerts, and send those papers to your supervisor, they'll appreciate it, you'll know what's going on, and you'll look good to your supervisor.
But: be careful not to spend too much time on this.
> you've likely had positive experiences with research as an undergraduate.
This guy is really really out of touch with reality. People that go get PhD’s are driven. I know many that had entirely shit experiences with undergrad research, horrible advisors, but new what they wanted and were darned determined to get it.
>Also, you've been a good student in school, scoring at the top of your class on exams and projects. And you've probably been praised throughout childhood for being a smart kid.
Umm, hmm. This is actually a problem. He’s looking for less diverse, vanilla, 1% kids. Sure, the entrance criteria are smart kids with good grades. Likely the best and most driven will be the scrappy ones from really odd backgrounds with few “projects” of official nature and even fewer actual undergrad research experiences. I hate people like this that walk the walk of “diversity” but then they go off and try to find the exact same as themselves, silver spoon kids of academics. If you look at the really successful people, they have some really interesting and truly diverse backgrounds. Just my .02
If you want to get a PHD, do it because you love learning, teaching, are extremely creative and have tremendous follow through. Are you a person that constantly thinks, “why am I here?” The. Get out and find a job, it’s not worth it.
I think the first paragraph should've said everybody's experience is different.
Doing a PhD is like joining a company. It's a job. Depending on the team, manager, department, project, your personality etc. your experience will wildly vary.
> UNDERGRAD VERSUS PH.D. RESEARCH (being very different)
Here's my counterexample: My undergrad research experience was much like my MS/PhD research experience. It even led to a paper.
Advice I would have liked: Don't do a PhD if you like taking classes and just want to extent that experience, as doing a PhD is very different from taking classes. Also, the classes you are forced to take probably suck.
Advice I received but ignored but is probably true: Doing a PhD actually decreases your chance of getting a job (especially in Computer Science)
Advice that I would have liked to receive, know now but still ignore: If you are in a foreign country and you don't know the language but you want to stay in the country after you graduate, learn the language! Don't spend all your time doing research thinking that you will somehow compensate for not fluently speaking the language just because you have more research achievements. (Disclaimer: I live in Japan, and most jobs, even academic jobs, require a sufficient level of Japanese language abilities)