[malrase]

Your supervisor is key. You have to like them, because they’ll be criticising your work for the next 3-4 years.

This sentence resonates with me particularly. I was encouraged to choose my PhD based on love of topic rather than supervisor. I had multiple topics I enjoyed, but ended up going with a worse supervisor and a topic I thought more interesting than a better supervisor and a slightly less enthralling topic.

After a year, I quit my program.

What I learnt: even though a topic might be slightly less interesting, the people you work with (and in particular, your supervisor) can make you love something more.

(NB: Obviously I wouldn't recommend choosing a topic you have no interest in and/or hate just because you get along with the supervisor!)

[xk_id]

How do you know whether your supervisor is a good fit? I have heard this advice, and similar stories, several times. But at the moment (I'm finishing my undergrad), it is not clear how the student is matched to a supervisor. For my dissertation, we were reandomly assigned teachers, and they insisted we cannot change them, etc.

Do you normally have time to chat with your potential supervisor, before deciding if you think it could work?

[xijuan]

I found it is a bit odd that you couldn't pick your supervisor. I think in most universities who you want to work with are up to you and the supervisors who wish to accept you. The best advice for finding a good supervisor is to ask current graduate students. Some supervisors are well known to be good supervisors among grad students whereas others are known for being super mean to graduate students.. Current grad students usually know who are good at supervising. Also ask about the style of supervising. Some supervisors are very hand off but you get little support from them. Others can help you a lot but may impose their ideas onto you.

[carterschonwald]

this times 1000x.

Had exactly that same grad school experience myself.

Lucked out after I left, and i'm doing even cooler stuff research wise in an industrial context! Building my own tech product that really require doing great Research and Dev both!