[pezzana]

Missing from this actionable survey is the skill of answering questions that haven't yet been answered in the secondary literature (books, reviews) or even primary literature (journals). This is, of course, what research and a PhD is all about in the end.

In my experience, this fact was the #1 reason why some peers dropped out of PhD programs. They joined expecting a continuation of undergraduate education, which consists largely of recapitulating what appears in the secondary literature. Instead, what they got in graduate school was the expectation that they would be producing the primary literature. That's a very different game.

It was a game these peers discovered they hated playing. Nothing in college can prepare you for the isolation of spending your time becoming the world's expert on a narrow technical topic. Your usual reinforcement mechanisms of approval from family and friends gives way to slight comprehension at best. Then there is all of the alone time doing research requires. But I suspect the hardest part of all is the seemingly endless lineup of dead ends and false hope. Not only is success not assured, you often have no idea whether the result will have any utility even if you succeed.

Then, just when you've gotten the hang of this finding answers game you discover that the real expectation is to be the one who formulates good questions. The kinds of questions that, although they will certainly involve dead ends, will ultimately pay off in some meaningful way. Very little in a bachelor's prepares you for doing this. It's a hard-won skill that comes from a round or two (or three or four) of months (or years) spent answering questions that nobody cares about. A lot hinges on your relationship with your advisor on this one.

The PhD isn't just a bachelors degree but harder. It's a completely different animal. The skills in this article are very useful toward that end. But there's a lot more to the story when it comes to skills for finding answers to those unanswered questions, and formulating worthwhile questions without answers.

The benefit of all of this work and discomfort is that you come away with the ability to answer worthwhile questions that haven't yet been answered. And that's a highly transferrable and applicable skill.

[tchalla]

> In my experience, this fact was the #1 reason why some peers dropped out of PhD programs. They joined expecting a continuation of undergraduate education, which consists largely of recapitulating what appears in the secondary literature. Instead, what they got in graduate school was the expectation that they would be producing the primary literature. That's a very different game.

Hit the nail on the head. I would like to add one point though - it's not just the unanswered questions, one sometimes doesn't even know which questions are unanswered.

Typically, up until a Ph.D - you are given a question and then asked for an answer which more often than not exists. Suddenly, in a Ph.D - not only do you not know the answer, you don't know the question too. The craft to come up with an important question, create a well-defined scope and then answer the question from different perspectives is the heart of a Ph.D program. The true skill is the ability to "learn to learn". The transferrable skill is to probe around for questions which are important, define them and then go ahead to answer them.

[tasogare]

From my experience failing my PhD it's not so simple. The head of my lab have a bunch of topics he want to (make other people under him) investigate, sometimes really precise. One of the PhD student literally got his thesis question handed down after an experimented researcher worked on it for 6 months. Other like me had to found one themselves. It's obvious that the first student got a head start of about a year. The irony is he is now in difficulty writing his thesis, despite having published the required number of papers, which is not too surprising since he didn't get the problematic by himself.

[rabbits77]

I recall the same things. I remember being bitterly envious of students that seemingly were given topics to examine while I had to spin my wheels for a good year or two coming up with my own ideas to explore.

Looking back my perspective is very different. First of all, my ideas generated a minor spike in publications for the lab all centered around my work. After I graduated I continued to advise new students to continue what I started.

I now think in these teems, which might sound cynical but simply reflect the vicious nature of academia.

    1. The students given topics, the ones I was jealous of, all kind of sucked. They were given topics to advance a short term goal and gtfo. It may seem paradoxical or cruel but a savvy advisor will maximize both short term and long term gains. I was a long term bet, the others were short term plays.

    2. I benefitted greatly from being forced to identify my own topics. This is the one skill that I use every single day. Every hour of every day. As time goes on the ability to evaluate ideas deeply and with some speed effectively defines what it is I do for a living. 

    3. Students given topics were cheated out of more valuable long term skills for the lab’s short term gains. This is not always universally true, of course. Some super stars really can crank through a deep serious topic quickly and continue on to generate novel research if their own. One such person may appear in a university department once every ten or twenty years, they are extraordinarily rare.

    4. I was well aware of the exploitive nature of grad school, did it anyway with a clear head for what I wanted to get out of it and my only disappointments came from when I giddily let my guard down and expected more than what I already realized would be forthcoming. A specific example, my advisor would use my conference paper acceptances to fund their own personal travel and vacations; I was not allowed to present my own first author papers. Silently tolerating that sort of bullshit, in part, allowed me to graduate.

[mcguire]

"A specific example, my advisor would use my conference paper acceptances to fund their own personal travel and vacations; I was not allowed to present my own first author papers."

That's really very bad. Learning to present to your peers is an important part of the process, as is getting your name and face out in the field.

[anonymousDan]

Wow, not being allowed to present your own first author paper is pretty bad - unless someone else contributed a lot of the work and was made second author + presented?

[rabbits77]

Not at all. The papers where I was first author were essentially 100% all my work. Second or third authors were advisor and a committee member that helped edit, closely read proofs and pseudocode. Certainly helpful and deserving of authorship but a steep drop off in contributions from first author (me) to them.

Advisor had family in Europe (I am in US), and would use these conferences to create extended family vacation plans.

Of course when I think back on this I am still pissed! But years have passed and I am doing well in my career. I am able to not get too fixated on it.

And yes, I agree with the people saying it was terrible and very wrong. In terms of all the possible terrible and very wrong things that can happen to a graduate student this is maybe, in the grand scheme of things, about mid-range maybe?

[jjdredd]

Wow, those who could find their own research topic were lucky. I've never seen anyone in my environment get that much freedom. The supervisor sets the problem and the student must solve it.

Edit: I was under the impression that even the postdocs are hired for a specific task.

[BeetleB]

> Suddenly, in a Ph.D - not only do you not know the answer, you don't know the question too.

Mirroring what tasogare said: There are a lot of research professors who will not give you the flexibility of finding the question. They often are paying you to be an RA, and will want you to work on their topics, not yours.

This may vary per discipline. In the circles I was in, this was the norm, though. Some professors were open to you choosing your own topic, but the "contract" was similar: If they are funding your research, then you should work on your own topic "on your own time".

[tchalla]

I know what you and tasogare refer to and I've seen it happen. I had to fight to change my thesis topic 3 times and also change advisors. Of course, it is not simple.

I wouldn't change anything from my initial comment though. Flexibility is not binary - it's a gray scale. If you have ZERO flexibility, you should accept the implication that such a Ph.D will be stripped of some valuable lessons. On the other hand, you can always decide to not do it and move to a different professor. You could also decide if the broad area is ok with you before you take an admission to a lab.

[yt-sdb]

I'd frame it this way: a good senior researcher knows when to give a junior researcher the question or not. First-year grad student? Pair them off with a postdoc for fast iterating on an established project. Fourth-year grad student? Push them by letting them flounder a bit and learn to find their own questions.

I agree that in practice, this simply varies a lot by discipline and advisor.

[mirker]

You missed the most fun option: pair them up with a problem that leads to nowhere!

[mcguire]

This is all completely true.

"Then, just when you've gotten the hang of this finding answers game you discover that the real expectation is to be the one who formulates good questions."

And that is where I personally failed.

On the other hand, there's that funny moment...

One of the things I've heard repeatedly from pilots is that first solo flight changes everything. Before that, you're just some human. Afterwards, you are some human who can fly. Everything is somehow different, although I've never seen anyone really successfully describe how. I suspect it's different for everyone. But then I'm not a pilot.

In your dissertation defense, someone whose knowledge and intelligence you respect immensely will ask a difficult question. When you answer that question confidently and to their satisfaction, the world is a different place. For one thing, you're no longer student and teacher; you are peers. But that's not all it is.

[lambdatronics]

Spot on! I had the same misconception, but it worked out OK b/c I was motivated at least in part by curiosity. When you're curious, you ask enough questions to get to the edge of knowledge and then pose a novel question. If you enjoy the coursework purely b/c you like having nice tidy answers to everything, being at the edge is uncomfortable & research isn't for you. OTOH, in my PhD coursework, the HW questions were almost always solved ones where we just had to reproduce the steps to get the answer that was included in the question formulation; this burned much of my curiosity out by the time I was done.

[jiggunjer]

Those dead ends are negative results. While not easily publishable they can form the bulk of a thesis. Many people get demotivated because they treat a thesis like a journal publication. I, for one, was glad I finally didn't need to sex up the language to convince some editor.

[lambdatronics]

> need to sex up the language to convince some editor.

I hated this pressure. I wrote up the core of my thesis as a manuscript for a second-tier journal, but my advisor though I had a shot at a first-tier publication. I disagreed, but I rewrote the paper anyway, and had to significantly rework/descope it. It ultimately wasn't accepted for the first-tier journal, so I rewrote it a second time for the original journal. The whole process was immensely frustrating (cat-herding coauthors, playing volleyball with editors/referees, trying to discern whether my concerns about overselling my results were legitimate issues of integrity vs. instances of imposter syndrome, ...).

I fell in love with the hard sciences because "reality must take precedence over public relations, for Nature cannot be fooled." [Richard Feynman] Finding out how much PR is actually involved was hugely disillusioning.

[mizzao]

> The benefit of all of this work and discomfort is that you come away with the ability to answer worthwhile questions that haven't yet been answered. And that's a highly transferrable and applicable skill.

This is why I have come to see that PhDs can in some cases make excellent founders. Source: CS PhD turned founder ;-)

[OccamsRazr]

And this makes everything else in the survey easier. Topic sentences and presentation skills are useful but most important is having something original and substantial to say. The rest follows and is easy by comparison.