This is great! The ending phase really gives you the same feeling of “screw everything else, I need to finish this paper” rushing to get the final stuff out.
By that time I had already grinded so long and lost so much hope that I HAD to slack off but couldn’t because every time I tried, my advisor caught me and reminded me of my progress. Repeat until I lost all hope!
This is great and exactly captures the PhD experience. Both in the simulator and in real life, I mostly survived until the end by slacking off frequently, and needed around 5 years.
Some highlights:
> INBOX: Based on the reviewers' comments, we regret to inform you that your manuscript has been REJECTED for publication. One of the reviewers pointed out that there is no comparison with a state-of-art method.
> You came up with a bunch of ideas. However, upon further searching, you found that they have already been done before.
> You found the missing piece during a shower. You develop one of your preliminary results into a major result.
> You found one of your ideas appears in a recently published paper. You can no longer work on it.
> Three years passed. You have witnessed many graduations. You began to worry about whether your can graduate on time.
> The simulation took a much longer time than you expected. The results are not available yet.
What was missing:
1.) Growing feeling of getting too old
2.) Growing family obligations (marriage, kids, trying to write a thesis at 3am with a crying baby next room)
3.) Questions asked by friends and relatives regarding progress
> 3.) Questions asked by friends and relatives regarding progress
This one in particular had me temporarily cut off contact with people who could not be bothered to remember that I had no interest in answering this question!
One dynamic I experienced that also isn't in the simulation: if you focus too much on classwork early on in order to pass your RPE, it can actually be hard to find an advisor. Classwork is basically dead-end work and the more you focus on it the less you have to show for yourself when trying to convince an advisor to work with you. Your goal should be to optimize for doing just well enough to pass your classwork.
Also, random catalysmic events, like in year 4 your advisor accepts a job at a different university in another state.
Every department I've worked in mandates 2 supervisors to mitigate against this, because it's reasonably common for people to move departments, go on sabbatical or just quit. Even with tenure, life happens and people need to leave their jobs. In theory, the department shouldn't allow advisors to take on more students if they're close to retirement though.
Turkish university system solved that problem by stipulating that thesis supervision and advice duties survive retirement unless the student explicitly requests otherwise, and end only if the student finishes or the supervisor/advisor is fired (not resigned!) or dead.
This also depends on the field. I think this is good advice in fields where grad students are primarily there to help with grant-funded research. In those fields, the course and prelim requirements are reasonable because professors need warm bodies doing work. Eg, CS.
It's less good advice in fields where grad student research output doesn't matter as much, and where students do more teaching instead. Those fields tend to make much more aggressive use of weed-out exams to ensure that they have enough young grad students to meet teaching demand but not so many older (>=3yr) grad students that they saturate advising capacity. Mathematics in particular comes to mind.
> Classwork is basically dead-end work and the more you focus on it the less you have to show for yourself when trying to convince an advisor to work with you.
nearly half of my year didn't get this and had to master out when we got to quals.
Are those different things? At departments where quals have high failure rates, it's really more of an annual layoff than anything else.
In many programs, the department aims to admit far more people than will pass the quals. They need the Calculus and Pre-calculus TAs but do not have the advising capacity.
Even if everyone gets a 95% on the quals, the majority will "fail" by necessity because the department simply does not have the advising capacity for the number of TAs they need. Of course, the department typically designs the quals to these needs either explicitly or implicitly.
This is usually at least implicitly understood by the faculty, who will navigate it when absolutely necessary. For example, I've seen it happen that if a professor really needs a student and vouches for/protects them (eg because the research is computational and the student came from 5 years at Google), then the student gets more goes at the plate on quals than is typical.
Depends on the school. My program requires you have an advisor before our version of quals. Our exam was also a research presentation followed by an oral exam where you had the pleasure of doing often a painful amount of algebra on a blackboard. It was also the culture that, while your advisor could ask you questions, they rarely would unless they were helpful (maybe a softball confidence booster or simple version of another question to get you started down the right track) or leading questions for you to talk about something not in your talk. People were largely concerned about being failed by the other examiners, not their advisor at that stage.
However, we were a small department and your acceptance was predicated on at least one faculty member wanting you to join their lab. Some people moved around within their first year, but the majority stayed with the first group they joined.
er, right, what other people call quals we called prelims. our quals were a presentation of early stage work, so when the deadlines rolled around, they didn't have an advisor, or hadn't been working with them long enough to have any results.
It's funny, my CS program in US you had to have an advisor by end of your first year and your quals were basically the thesis proposal you did with your advisor.
This would be incredibly bad advice in half of Physics and most of Math. An adviser would simply not trust a graduate student with middling grades to be competent enough to work with.
I think this depends on the field. For example in CS, my advisor straight up told me multiple times to stop worry about class work. His exact statement is that "There isn't anything more that you will learn in classes that you won't learn in greater detail doing research".
His logic is that when you are doing research, you are pushing the envelope into new territory that can't be taught in a classroom. When you are in a classroom you are learning old material that is already well-known and established.
This is very true in CS. But far from true in Math and Physics where there probably is a lot of advanced learning available in classes. The few classes I had that he actually endorsed being "worth your time" were Math classes focused around encryption (of which I took 3 different ones).
But my advisor was unique because he was 100% there for the research. He only taught because the university forced him to. He lived and breathed research and that was the only reason he was in academia. He was truly passionate and worked 10+ hours a day on research, but thats why he was there. He had a very low opinion of classroom teaching.
IME it's more that the advisor doesn't trust the student to make it through the annual layoffs (quals culling), and only wants to invest in people who they know will be around long-term.
At least in the poor (and honestly mostly useless) parts of Mathematics. Maybe Physics is less poor.
(Fortunately I was in CS, where the research output is actually needed by society and usually not pure masturbation, so the attitude toward coursework was "do well at what you need, enjoy what you want, and ignore what you don't need or want"
Most universities do not have "quals culling" and most universities that have "quals" (not that common anymore) do not use them for culling. There are a few notorious exceptions though.
As a counterpoint, I knew a physics prof that would drop any grad students who got above a B average, since it meant the student had bad time management. Bad time management being defined by spending more tine outside the lab than strictly necessary.
Then again, knowing that that @£&$€¥ would drop you might make it a good plan.
This emphasizes how much you need to know before you even get into a program. If you don't know that much, you NEED those classes to just catch up to your peers as to understanding what the world even says about various things at a foundational level. The weight is so much easier to carry if you go in with a certain level of knowledge so that you can slack in classes if you need to rest.
This. Most grad school classes are poorly taught and the professors indifferent or discouraging to actually helping you learn. PhD students are assumed to be capable of learning these things on their own or already knowing them. If you are encountering things for the first time, you'll likely be behind.
In contrast, if you come in mostly ready to go and these classes are just refreshers, you can spend time in that class working on actual research and impressing the prof as well as not panicking if/when you realize you don't understand what's going on.
I had picked my advisor at the start of my PhD. I also had 2 backups. My pick was on sabbatical my first year. He and I agreed I'd load up on the required classes that year.
He e-mailed me right before the year ended saying he had changed his mind and didn't want any more grad students, basically dumping me.
Right around the same time, my first backup decided to retire.
My second backup passed away.
I was left no longer making "sufficient progress" and no path to do so, losing my financial aid.
Thank you! I didn't have many options at the time, having a small family to provide for, so I went back to work as a software developer. The career has been good. I regret not being able to be a professor and devote more time to research, but I also try to make opportunities for those kinds activities in other ways.
Generally, the focus here should be on:
1) Not bombing any classes (i.e. A/A- in all, maybe a B+ in one; a B or below is failing)
2) Doing very good work and trying to write an original paper for professors that you want to work with while doing just enough to get by in other classes [this is in part how you figure out who you want to work with]
3) Being good enough with the literature to pass the comprehensive exams (or, as another comment points out, have some kind of protection from a sponsor; it is not uncommon to have profs use comps as a chance to take out students they don't like for various reasons, even as small as "they do X field, which I don't like" or "they work with Y, who really gets on my nerves).
Of course there's plenty of additional ways to derail this as well, including advisor moving, advisor getting into a fight with the rest of the department, advisor giving poor advice, advisor deciding that they don't like you, etc.
It really makes you wonder if the university should just have a mechanism to "fire" a grad student rather than pretending that these events aren't simply a mechanism to "fire" someone because they didn't pass X hurdle.
If the advisors can vouch for, or strike a student regardless of their qual performance - then why not simply have an end of year performance review?
Most depts do have some kind of official review, but it's more of a formality. I think they're also concerned about how students would react if they suggest that academia isn't for them directly. So instead they resort to more passive-aggressive or arbitrary measures.
On the other hand, not all departments are good fits with students and there's a very wide asymmetry in information between many new students and programs, even if you "do your research" beforehand, given just how specialized these disciplines are at a high level. It would be nice if transferring programs was made easier and if more departments would just agree to help students "master out" and look for jobs rather than discard them like roadkill.
For many schools that’s basically their qualifying exam. Normally multiple professors are involved so that your advisor can’t fire you unjustly, but if you don’t demonstrate you are up to standards on the exam and your advisor doesn’t go to bat for you in their deliberations, you will fail. Normally you have a chance to retake the exam, but two strikes and you’re out.
In my program we had annual committee reviews as well as a review submitted to the program chair and graduate school of the student by their advisor (with the student providing both a self-review and review of their advisor). Ultimately, unless you’re in a sub-field with many faculty, it is hard to get an accurate evaluation from three professors. My other committee members could understand the big picture of my work but they weren’t experts on the specifics, and they were the best equipped faculty in the department to be on my committee besides my advisor. The goal is to make sure that there is a paper trail and multiple professors aware of your progress (or lack thereof), so your advisor can’t just give you the boot for something tertiary like not watching his dog during a holiday weekend.
Professors are aware of their problematic peers in the department. Even if they can’t fire them outright (tenure has pros and cons), they can steer students away from them to more supportive professors (or give you a hint that maybe you should consider a different school during your visit day). Our program chair was very good at helping relocating students who initially started in the lab of one or two bad actors.
Ironically universities have already found a way to defeat that behavior, too:
At some places, advisors now also consider a list of students ranked by grades. I have seen advisorship being offered only to the top student in the Professor's class.
If you lack ex-ante information how and where your grade rank matters in a year, this just adds a fun new challenge to the first year PhD!
That depends on your department and your field. Even though, on paper, I had a year to pick an advisor, I worked with mine from day 1 (or day -n since he first sent me research papers and discussed ways to get started on my project before my first semester even started). My department was fairly small, though, so you getting admitted meant at least professor specifically was interested in you joining. Professors at larger schools will put on their website not to even contact them unless you have been admitted, and their departments are large enough that only a portion of faculty members are involved in admission decisions. In that case, your advisor of choice may not know about you at all before you enroll and thus you need to fight for a spot in their group among your cohort.
> You found one of your ideas appears in a recently published paper. You can no longer work on it.
This is one of the things I thought of right away when ChatGPT got released last year. "God, there's probably so many PhD candidates right now in NLP feeling despair like all their work was pointless ...as if million of voices cried out in terror and were suddenly silenced."
It's hard in the moment to know whether what you're working on has any utility. So just do your best and keep chugging!
And this attitude, my friends, is the reason why so much software out there is so bad.
We need more of a math mindset when developing software. What can we be sure about, what are the invariants, what can we prove? There is so much crap out there that somebody lacking understanding just tried to wing, and I'm constantly ashamed of it.
Number theory had no applications for centuries. Now, cryptography is based on it and the modern internet would be unthinkable without.
Foundational research does often not provide immediate applications. Still, if we don't do it, out understanding of the world is lacking and it hurts us later down the road.
While there certainly exists math for the sake of math, there is a trickle down effect that is quite real (there’s also a trickle up effect that is real but that’s unrelated). Someone does some math for the sake of math. Later on, someone who is slightly more applied sees a link between that math and a more applied problem they’re working on. If the idea is truly useful, it propagates down all the way to application-focused practitioners. Researchers exist on a spectrum, generally, between pure theory and pure application.
Math has no application until you find an application for it. Differential equations are just equations until you pair them with physics. Formal logic is just an abstract discussion of human reasoning until you build a circuit, etc.
One wonders if trickle down mathematics is any more efficient than trickle down economics. It seems like we might be better off not funding pure math, as forcing function to coerce those minds to work on more applied problems directly, instead of relying on this random serendipity.
It seems like I might be better off picking the winning lottery numbers directly instead of relying on the random serendipity of guessing them and most of the time being wrong.
Its sort of a mix of a lot of small things - 1) The coming conferences will be flooded with LLM analysis, so the space will be heavily saturated and more difficult to find a significant contribution; 2) LLMs are a new model that you might need to include in your analysis, which means learning about and becoming familiar with them; 3) your work might get overshadowed because its now obsolete in the land of LLMs
A slight equivalent I can think about would be the emergence of neural networks. When I was working on my Masters on face recognition, neural networks were not the major force they are now. Facial landmarks used a combination of haar features and edge detection. These methods weren't outright abandoned, but if NNs had taken off during my research, then I would have had to restart my work.
I met someone recently who finished their PhD in computer vision related work a couple years ago and she said all of her specialization now felt useless, but that her PhD was still useful for understanding the fundamentals for a job she now has but does absolutely nothing with her research experience.
HmThat very heavily depends on the specialization. E.g, you did image processing, basically useless now, you did GANs, diffusion models took over. It's like that for probably most phds but the research skills, writing skills etc are with you forever.
Tried that, the university considered lecturers to be disposable if there was a chance to replace them with a tenure-track who could get grants, and told me I could come back if I got a PhD.
If you want to have a tenure track "professor" position focused on teaching in a top-tier university, you need to be a great researcher as teaching skills are not considered much -- you just get to decide to focus on teaching after you get tenure. Thankfully, many universities (even the prestigious ones) are now starting to hire more semi-permanent teaching-focused staff (and some even use the title professor for these). You do not get as much independence in such a position, especially if you want to make a class for more senior students, but it is a good middle ground. Or you can be a professor at a school that is not in the rat-race to be "top-tier research institution" - you still need to have some small research output (but that is actually an awesome way to introduce a couple of undergrads a year to research) and you get to focus on making awesome classes (of course, there is still the expectation that you have a PhD to apply for these positions, but at these places your teaching experience is actually taken seriously in the hiring process).
As to what to do during your PhD: find an advisor that is happy to have one or two students focused on teaching and outreach (they would like to have that because when applying for grants it makes it easier for them to explain how they have broader impact, pointing to your work).
> If you want to have a tenure track "professor" position focused on teaching in a top-tier university, you need to be a great researcher as teaching skills are not considered much
Except that UC Berkeley most definitely is “in the rat-race to be "top-tier research institution" (probably more accurately described as one of the rat-race winners in many fields) - it’s definitely important that top-tier research institutions also provide quality teaching.
We do agree on this. My second sentence is exactly what you just said:
> Thankfully, many universities (even the prestigious ones) are now starting to hire more semi-permanent teaching-focused staff (and some even use the title professor for these).
In academia, your Resume/CV is basically a list of what you have published.
Even if you are an awesome teacher, you are going to be required to continue publishing a minimum amount every few years and you will be hired based on what you published.
Sorry, but that's just academia. If you want to teach without doing research, then maybe look at Community Colleges, High Schools, or getting a job at a corporate job and being an Adjunct Professor for 1-2 classes a semester.
There are also smaller (typically private) colleges and universities that heavily focus on their undergrad programs. The Jesuits seem to lean into this style with both Santa Clara U. in the Bay Area and Loyola Marymount U. in LA falling into this pile. Research at these institutions definitely ends up taking a secondary role.
During the PhD, I was a TA and instructor on record for several classes. Schools may have some form of mentor teaching assistantship that lets you get experience teaching while in the program. I think I taught ~6 courses by the time I graduated.
It can also help to position yourself in the "education" research space for your field. There is a strong CS education research space, so you can incorporate your classroom as your "lab", though you'll want to study up on Cognitive Sciences to ensure your findings support current literature. My publication count is much lower than my peers, but I was still able to receive several offers for teaching faculty positions.
Teaching faculty positions are available, though not in as much demand as traditional research oriented profs. However, I know at least in CS there are several universities looking for them. Likewise, by situating yourself in the education space, you can land a research prof position while still focusing on education. If you get funding, then you can buy out course obligations so you can specialize in teaching a single class.
A teaching university and not a research university. You can / will still do some research but your job is teaching students not doing research. The pay is generally better, but of course, you will have to actually teach a lot and have a lot of office hours. Maybe once every few years you can work out a research semester. The initial pay is better but less so the opportunities for advancement as you won't be publishing much. That makes it harder to differentiate on anything other than time.
I'm not sure the salaries are better. Most R1s are now offering 90-120k starting in my field, but regional teaching Us start around 50-60k, with liberal arts colleges in the 50-80k range.
The point about the lack of opportunities for advancement/moving due to the course preps and teaching taking up your time is very true. While your friends at R1s are on pre-tenure sabbaticals, getting course buyouts, and teaching a nice grad seminar for a semester, you might be doing 3-4 new preps a year and likely getting piled with service work.
I would also add that the hope trajectory is quite right in the simulation. You really should start quite high and it starts dropping until your first conference, but then towards the middle of the PhD it gets very low. It only really goes up again when the end is in sight.
Yeah that’s how it went for me. My confidence was very high until I finished my qualifying exam and just had to sit there trying to solve a novel problem for over a year. Eventually I learned enough through trial and error that I could work productively in that space, but doing well in classes often does not translate to immediate research success and there were dark days and many therapy appointments in the middle.
The hope level can be lowest (and cynicism maximal) late and at then end when one is the senior student and the major (only?) contributor in the lab or key project with no one coming up behind. A weak thesis committee won’t stand up to the PI to drop the cycle of “just address this loose end” tasks. Meanwhile relatively non- or lowly- productive fellow grad students in the program are given quick PhDs in other labs because the labs are underfunded and those PIs don’t want the burden of retaining or the stigma of failing those students, and …
Wow—-pardon the rant I’m having a flashback. I’m going to go exercise my anxiety away,
I'm doing it right now and I kind of envy my colleagues that are doing normal work. There are times when I enjoy the ability to focus on things that really interest me, but the paper writing and publishing processes really suck. Also, the random stuff from the university that I have to jump over sucks fun out of the process, for no gain to anyone.
Ah yes, the coup de grâce. Or you do find employment and then find that your major result paper is being challenged and have to submit a retraction. The fun is endless.
That was...a lot more straightforward than I thought.
First try: Year 4 Month 5
Second try: a lot more things went wrong. Year 5 Month 11.
Third try: Year 5 Month 11.
I just followed these rules:
- Study for the qualifying exam until I'm "very confident"
- If I have no ideas, read papers
- If I have an idea, work on developing it. If I have a preliminary result, work on developing it. If I have a major result, conduct experiments etc...if I have a rejected paper, revise and resubmit. Prioritizing whichever option gets me closer to an accepted paper (because presumably the ideas get outdated quickly)
- Whenever I get the "ask my advisor for a break?" say yes. Whenever I get "I am tired" and no "ask my advisor", "Slack Off" for one month.
Fortunately I got no abusive advisor, rejected papers usually end up getting accepted later, no extreme life circumstances or cut funding. But my computer crashed way more often than I'd expect, especially since backups are so common nowadays.
Every time I tried the game (using OP's described steps) I "lost all hope and quit". I thought that was the joke / funny commentary: that all paths through the game involve you ending up losing. Didn't know there was a path to actually win the game until reading the HN comments. So I kept trying it over and over doing the same steps and finally won once.
I guess that's an even funnier commentary on how it's pretty much entirely luck based.
A bit disappointing it doesn't go into other details like being a women. As most PhD candidates are in their mid 20s and often by the end of the beginning of their academic careers they immediately have to decide to either have a family or pursue academia.
Didn't realise at the time but you have made me appreciate that that was one good thing about the British system - I was 24 when I submitted my PhD which is fairly typical so still time before such decisions (it was another 11 years before I discovered the challenges of combining a career with motherhood!)
Reducing a PhD program to 4 years can help but in many fields this is still a problem. For many fields and form of absence of leave can be the end of your academic career. If I recall correctly the average PhD graduate is still 26-27 years old. Which doesn't help much too much even though the average US grad is 31. For many field though this would help a lot. In tech and stem not so much. I remember some of the female faculty I knew telling me exactly how and why so many women end up filtered out. Simply because they wanted to have a child and by the time they came back their research was outdated and their works published by others. They themselves never having children. It's why many women often leave academia for industry jobs.
I personally believe how we conduct research and academia is outdated and does not allow for the proper inclusion of women. And does not allow men to be proper fathers. Sorry for the long talk.
Edit: Got the average age for PhDs in UK wrong it's mid 30s. Even if they started right at the age of 22 it's a wall they will face almost immediately.
In my opinion, a first step is to equalize maternity and paternity leave. It should be equally disrupting for men to have children as for women (from the perspective of an employer). I like the Swedish implementation of this model, where partners get 480 days of leave per child which they are free to divide among themselves, with a minimum of 90 days for either.
That's a ridiculous amount of leave. A man does not need much time outside abnormal circumstances for paternity leave. I would almost never work, as my wife and I have a baby every year or two.
> I was 24 when I submitted my PhD which is fairly typical
Good for you - that's very fast. According to [1] the median age for starting a PhD in the UK is 24 to 25 for full-time students. So you actually graduated around when the typical student starts.
I followed this strategy + parallelized developing preliminary ideas and wrapped up in 3 years, 11 months with 100/100 Hope. If it only it was that easy! Haha
There is also a hidden variable about being tired too, and if you don't slack off you waste months in forced situations with asking for breaks with the supervisor or burning out.
I basically followed this approach but my paper kept getting rejected from the conference which really put a ceiling on my hope; ended up quitting the PhD after just under 5 years.
Second try the conference paper got accepted right away. Advisor even asked me if I needed a break after I'd had some success (never happened on the first run) and was getting tired. Wrote my thesis in 5.5 years.
Also just pictures of them enjoying life in the city of their choice, the families they’ve started, the vacations they go on, the fun hobbies they have time for, etc.
Some time ago I watched this movie: https://www.imdb.com/title/tt0416675/ (Dark matter). I don't know how accurately captures the essence of getting a PhD, maybe someone who went through it can expand on that, but I've always wondered if it's actually how is depicted in the film.
This film is a fictionalized depiction of the real-life 1991 University of Iowa shooting. A PhD student murdered his advisors, VP of Academic Affairs, and a fellow student for being passed over for a prestigious prize. There were other geopolitical influences on this event that I'd say make it unique.
This sounds less stressful than an average half a decade working at my current position. It is actually motivating. Whether or not that was the desired outcome, I don't know. At least I get to travel to half decent academic conferences and not large vendor marketing conferences.
I'm interested in how specific this PhD experience is to the US, certain subjects or recent times. My own experience doing a maths PhD in the UK in the mid-2000s was not like this at all (but had a different set of challenges for sure).
I'm a chilean maths grad student and save for the qualifying exam, it's quite accurate. So much so that I think I made a mistake clicking on this because as it progressed I started feeling dizzy. Other commenters here also have their relatable experiences, which doesn't make me feel so bad.
I'm about to wrap up my PhD in a few months here in the US, I find that while it's kind of close, it's a bit on the cynical side, as is most of the HN discussion about PhDs.
Yes, my advisor emphasizes papers a lot, but there aren't any requirements for number of papers for graduation. While there are extremely busy periods of forgoing sleep to work (eg right before a major deadline), my advisor also constantly reminds us to take breaks and enjoy life. There was also the anxiety about graduating on time, but that too was sorted out by just having a meeting with my advisor and understanding how things work.
On the other hand, the situation with the qualifying exam was the opposite, I had to constantly remind my advisor that I needed to get that done. It involved a 50 page report on the current status of my research and a thesis defense style presentation to my committee, so that was a bit of a challenge to make time for between normal research. Passing it didn't feel like much of a challenge, just meeting the 50 page requirement did. I had enough data, but it was still a lot of writing.
>Professors demand you do exactly what they want for your thesis.
Also, what qualifying exam? It seems like as long as I was worker for my professors, they couldn't give a crap. (Although I was quite credentialed, so maybe they didn't care)
If I do get a PhD, it will be on a topic I want. So far, I have done that better independently and have gotten a bunch of press on the topic without needing academia.
That's what I did (~17 years ago). I was registered as MPhil then transferred to PhD after submitting a mini-thesis after about 2 years. In the end the mini-thesis actually contained all the most significant results that formed the PhD thesis and the latter just explored some applications.
This is not exactly how it works in STEM, at least not around me. Ideas tend to come from working on real-world projects, which then shows the lack of understanding and need for research. The project forks to do the research and merge back to implement the findings. Thereafter, someone on the team will put it into a cohesive academic format, and use it for a PhD. Of course there is reading papers and such, but it is not the source of the idea.
edit: I am also curious, how many really stop research because a similar or tangential topic was explored? "There can be only one"?!
I'd say this is where more engineering approachable ideas come from or ideas that might be considered translational (or near translational). If you work near the bridge of theory and application, this is probably more the case. A lot of funding agencies are pressuring even theoretical or basic research to look at translational application anymore though (capitalism is all about immediate ROI) so I'd say this is increasingly shifting towards many research domains but in the past, there were a lot of pure theoretical areas you could work on where there are plenty of unexplored ideas you could build out and get paid to do the work.
Application is certainly a great driver though because you have a demand signal to look at vs throwing darts at the board in work that may never manifest to anything solid.
> This is not exactly how it works in STEM, at least not around me
You might want to take a different approach to reading papers then. No paper ever concludes by saying "yeah our method is perfect and no further work is needed" [1]. Instead, every solution has its quirks and questions which need to be explored further. Maybe their method has limitations which make it unusable for your applications, maybe they make somewhat faulty assumptions that don't always hold, maybe they wrongly ignore some technique. Seeing how other people approach a problem can often give you inspiration for how to take it another way.
It's been mostly like this for a long time, but it is slowly changing. Open data repositories and scientific software libraries/products are beginning to count more and more (at least many of us are pushing for this). It also depends on the target career past graduation. Papers matter a lot for tenure-track positions, and much less for science support (scientific software developers, data engineers, lab managers etc.) in academia, or most jobs in the industry.
The 3 paper requirement in the game is also not a formal requirement in most universities--it's more of an implied requirement by individual PhD advisors. FWIW, my first lead-author paper I published a year past my PhD. During my PhD, I produced two relatively large scientific software applications (one open and one closed source) and a few open datasets. I'm now 8 years past my PhD and relatively successful in my field, 90th or so percentile based on common metrics--papers, citations, and funds raised.
Bottom line, papers are important but not the only thing that counts. Outside of tenure-track careers where they are crucial, it's possible to establish yourself as a scientist and be respected by your peers by publishing software and data.
In what fields is 3 papers expected? In my field, Psychology, 3 first-author papers sounds like a reasonable lower bound, but that seems like it would be a lot to expect out of Biologists or hard scientists.
Current 4th year in Biology. Generally, 1 paper is expected, but it is not strictly necessary to graduate. Highly doubt that the PI will let you go without finishing your project though.
A very important part of the work of the advisor is to pick a subject that is not been researched by other group, so you can work on it without the fear of been sniped. It not foolproof, but if three or four ideas get sniped, it's probably better to kindly consider an advisor change.
As a current 4th year graduate student, I opened the link and found the choice to accept or decline the admission offer, I don't know why I clicked decline without thinking.
Your prize is always wondering if you should have done it. I know people that say one of their biggest regrets was electing to pass on the PhD path. The grass, unfortunately, is always greener.
"You decline the PhD offer and end up happy with an intellectually satisfying six-figure role in tech. You are surprised when you learn that your older colleague with a PhD earns just as much as you do."
284 comments
[ 5.3 ms ] story [ 257 ms ] threadThanks for the flashbacks. At least we didn't have any qualifying exams.
Some highlights:
> INBOX: Based on the reviewers' comments, we regret to inform you that your manuscript has been REJECTED for publication. One of the reviewers pointed out that there is no comparison with a state-of-art method.
> You came up with a bunch of ideas. However, upon further searching, you found that they have already been done before.
> You found the missing piece during a shower. You develop one of your preliminary results into a major result.
> You found one of your ideas appears in a recently published paper. You can no longer work on it.
> Three years passed. You have witnessed many graduations. You began to worry about whether your can graduate on time.
> The simulation took a much longer time than you expected. The results are not available yet.
What was missing:
1.) Growing feeling of getting too old
2.) Growing family obligations (marriage, kids, trying to write a thesis at 3am with a crying baby next room)
3.) Questions asked by friends and relatives regarding progress
4.) Teaching obligations
This one in particular had me temporarily cut off contact with people who could not be bothered to remember that I had no interest in answering this question!
Also, random catalysmic events, like in year 4 your advisor accepts a job at a different university in another state.
It's less good advice in fields where grad student research output doesn't matter as much, and where students do more teaching instead. Those fields tend to make much more aggressive use of weed-out exams to ensure that they have enough young grad students to meet teaching demand but not so many older (>=3yr) grad students that they saturate advising capacity. Mathematics in particular comes to mind.
nearly half of my year didn't get this and had to master out when we got to quals.
Or failed quals?
In many programs, the department aims to admit far more people than will pass the quals. They need the Calculus and Pre-calculus TAs but do not have the advising capacity.
Even if everyone gets a 95% on the quals, the majority will "fail" by necessity because the department simply does not have the advising capacity for the number of TAs they need. Of course, the department typically designs the quals to these needs either explicitly or implicitly.
This is usually at least implicitly understood by the faculty, who will navigate it when absolutely necessary. For example, I've seen it happen that if a professor really needs a student and vouches for/protects them (eg because the research is computational and the student came from 5 years at Google), then the student gets more goes at the plate on quals than is typical.
However, we were a small department and your acceptance was predicated on at least one faculty member wanting you to join their lab. Some people moved around within their first year, but the majority stayed with the first group they joined.
His logic is that when you are doing research, you are pushing the envelope into new territory that can't be taught in a classroom. When you are in a classroom you are learning old material that is already well-known and established.
This is very true in CS. But far from true in Math and Physics where there probably is a lot of advanced learning available in classes. The few classes I had that he actually endorsed being "worth your time" were Math classes focused around encryption (of which I took 3 different ones).
But my advisor was unique because he was 100% there for the research. He only taught because the university forced him to. He lived and breathed research and that was the only reason he was in academia. He was truly passionate and worked 10+ hours a day on research, but thats why he was there. He had a very low opinion of classroom teaching.
If they do not make the grading lenient - it’s for a reason.
At least in the poor (and honestly mostly useless) parts of Mathematics. Maybe Physics is less poor.
(Fortunately I was in CS, where the research output is actually needed by society and usually not pure masturbation, so the attitude toward coursework was "do well at what you need, enjoy what you want, and ignore what you don't need or want"
Then again, knowing that that @£&$€¥ would drop you might make it a good plan.
In contrast, if you come in mostly ready to go and these classes are just refreshers, you can spend time in that class working on actual research and impressing the prof as well as not panicking if/when you realize you don't understand what's going on.
He e-mailed me right before the year ended saying he had changed his mind and didn't want any more grad students, basically dumping me.
Right around the same time, my first backup decided to retire.
My second backup passed away.
I was left no longer making "sufficient progress" and no path to do so, losing my financial aid.
Of course there's plenty of additional ways to derail this as well, including advisor moving, advisor getting into a fight with the rest of the department, advisor giving poor advice, advisor deciding that they don't like you, etc.
If the advisors can vouch for, or strike a student regardless of their qual performance - then why not simply have an end of year performance review?
On the other hand, not all departments are good fits with students and there's a very wide asymmetry in information between many new students and programs, even if you "do your research" beforehand, given just how specialized these disciplines are at a high level. It would be nice if transferring programs was made easier and if more departments would just agree to help students "master out" and look for jobs rather than discard them like roadkill.
In my program we had annual committee reviews as well as a review submitted to the program chair and graduate school of the student by their advisor (with the student providing both a self-review and review of their advisor). Ultimately, unless you’re in a sub-field with many faculty, it is hard to get an accurate evaluation from three professors. My other committee members could understand the big picture of my work but they weren’t experts on the specifics, and they were the best equipped faculty in the department to be on my committee besides my advisor. The goal is to make sure that there is a paper trail and multiple professors aware of your progress (or lack thereof), so your advisor can’t just give you the boot for something tertiary like not watching his dog during a holiday weekend.
Professors are aware of their problematic peers in the department. Even if they can’t fire them outright (tenure has pros and cons), they can steer students away from them to more supportive professors (or give you a hint that maybe you should consider a different school during your visit day). Our program chair was very good at helping relocating students who initially started in the lab of one or two bad actors.
At some places, advisors now also consider a list of students ranked by grades. I have seen advisorship being offered only to the top student in the Professor's class.
If you lack ex-ante information how and where your grade rank matters in a year, this just adds a fun new challenge to the first year PhD!
This is one of the things I thought of right away when ChatGPT got released last year. "God, there's probably so many PhD candidates right now in NLP feeling despair like all their work was pointless ...as if million of voices cried out in terror and were suddenly silenced."
It's hard in the moment to know whether what you're working on has any utility. So just do your best and keep chugging!
PhD is granted for novelty, not practicality.
And this attitude, my friends, is the reason why so much software out there is so bad.
We need more of a math mindset when developing software. What can we be sure about, what are the invariants, what can we prove? There is so much crap out there that somebody lacking understanding just tried to wing, and I'm constantly ashamed of it.
Computer science is applied math.
Foundational research does often not provide immediate applications. Still, if we don't do it, out understanding of the world is lacking and it hurts us later down the road.
Math has no application until you find an application for it. Differential equations are just equations until you pair them with physics. Formal logic is just an abstract discussion of human reasoning until you build a circuit, etc.
A slight equivalent I can think about would be the emergence of neural networks. When I was working on my Masters on face recognition, neural networks were not the major force they are now. Facial landmarks used a combination of haar features and edge detection. These methods weren't outright abandoned, but if NNs had taken off during my research, then I would have had to restart my work.
any advice for people aiming for teaching instead of all the publishing stuff?
As to what to do during your PhD: find an advisor that is happy to have one or two students focused on teaching and outreach (they would like to have that because when applying for grants it makes it easier for them to explain how they have broader impact, pointing to your work).
Not completely true. UC Berkeley at least has tenure-track lecturers, now apparently mostly referred to as "Teaching Professors" (https://apo.berkeley.edu/sites/default/files/teaching_profes...)
A random (old) job post for this: https://gsso.ce.gatech.edu/2022/01/12/tenure-track-teaching-...
> Thankfully, many universities (even the prestigious ones) are now starting to hire more semi-permanent teaching-focused staff (and some even use the title professor for these).
Even if you are an awesome teacher, you are going to be required to continue publishing a minimum amount every few years and you will be hired based on what you published.
Sorry, but that's just academia. If you want to teach without doing research, then maybe look at Community Colleges, High Schools, or getting a job at a corporate job and being an Adjunct Professor for 1-2 classes a semester.
During the PhD, I was a TA and instructor on record for several classes. Schools may have some form of mentor teaching assistantship that lets you get experience teaching while in the program. I think I taught ~6 courses by the time I graduated.
It can also help to position yourself in the "education" research space for your field. There is a strong CS education research space, so you can incorporate your classroom as your "lab", though you'll want to study up on Cognitive Sciences to ensure your findings support current literature. My publication count is much lower than my peers, but I was still able to receive several offers for teaching faculty positions.
Teaching faculty positions are available, though not in as much demand as traditional research oriented profs. However, I know at least in CS there are several universities looking for them. Likewise, by situating yourself in the education space, you can land a research prof position while still focusing on education. If you get funding, then you can buy out course obligations so you can specialize in teaching a single class.
The point about the lack of opportunities for advancement/moving due to the course preps and teaching taking up your time is very true. While your friends at R1s are on pre-tenure sabbaticals, getting course buyouts, and teaching a nice grad seminar for a semester, you might be doing 3-4 new preps a year and likely getting piled with service work.
Oh ... I felt that one :-/
Then give you the option a year later to congratulate him on the startup’s multi-billion exit.
In the game it pretty much continously went up.
n) Your old college friends have secured their material needs while you barely make rent.
n) a PhD student that joined the program after you just surpassed your number of publications.
n) your thesis supervisor just bumped you off primary author to contributor in your own paper.
I mean, woah. Even that's a little too far for phd-land
First try: Year 4 Month 5
Second try: a lot more things went wrong. Year 5 Month 11.
Third try: Year 5 Month 11.
I just followed these rules:
- Study for the qualifying exam until I'm "very confident"
- If I have no ideas, read papers
- If I have an idea, work on developing it. If I have a preliminary result, work on developing it. If I have a major result, conduct experiments etc...if I have a rejected paper, revise and resubmit. Prioritizing whichever option gets me closer to an accepted paper (because presumably the ideas get outdated quickly)
- Whenever I get the "ask my advisor for a break?" say yes. Whenever I get "I am tired" and no "ask my advisor", "Slack Off" for one month.
Fortunately I got no abusive advisor, rejected papers usually end up getting accepted later, no extreme life circumstances or cut funding. But my computer crashed way more often than I'd expect, especially since backups are so common nowadays.
I guess that's an even funnier commentary on how it's pretty much entirely luck based.
I wonder if there are other issues you can get e.g. if you slack off too much but your hope is still low, it just stops working.
I thought that was a trap, and I was surprised my initial strategy of "say no but then slack off" didn't work
I personally believe how we conduct research and academia is outdated and does not allow for the proper inclusion of women. And does not allow men to be proper fathers. Sorry for the long talk.
Edit: Got the average age for PhDs in UK wrong it's mid 30s. Even if they started right at the age of 22 it's a wall they will face almost immediately.
Good for you - that's very fast. According to [1] the median age for starting a PhD in the UK is 24 to 25 for full-time students. So you actually graduated around when the typical student starts.
[1] https://www.hepi.ac.uk/wp-content/uploads/2020/06/PhD-Life_T...
Second try the conference paper got accepted right away. Advisor even asked me if I needed a break after I'd had some success (never happened on the first run) and was getting tired. Wrote my thesis in 5.5 years.
https://en.wikipedia.org/wiki/University_of_Iowa_shooting
Yes, my advisor emphasizes papers a lot, but there aren't any requirements for number of papers for graduation. While there are extremely busy periods of forgoing sleep to work (eg right before a major deadline), my advisor also constantly reminds us to take breaks and enjoy life. There was also the anxiety about graduating on time, but that too was sorted out by just having a meeting with my advisor and understanding how things work.
On the other hand, the situation with the qualifying exam was the opposite, I had to constantly remind my advisor that I needed to get that done. It involved a 50 page report on the current status of my research and a thesis defense style presentation to my committee, so that was a bit of a challenge to make time for between normal research. Passing it didn't feel like much of a challenge, just meeting the 50 page requirement did. I had enough data, but it was still a lot of writing.
>Professors demand you do exactly what they want for your thesis.
Also, what qualifying exam? It seems like as long as I was worker for my professors, they couldn't give a crap. (Although I was quite credentialed, so maybe they didn't care)
If I do get a PhD, it will be on a topic I want. So far, I have done that better independently and have gotten a bunch of press on the topic without needing academia.
This is not exactly how it works in STEM, at least not around me. Ideas tend to come from working on real-world projects, which then shows the lack of understanding and need for research. The project forks to do the research and merge back to implement the findings. Thereafter, someone on the team will put it into a cohesive academic format, and use it for a PhD. Of course there is reading papers and such, but it is not the source of the idea.
edit: I am also curious, how many really stop research because a similar or tangential topic was explored? "There can be only one"?!
that said, the game is fun! Thank you.
Application is certainly a great driver though because you have a demand signal to look at vs throwing darts at the board in work that may never manifest to anything solid.
You might want to take a different approach to reading papers then. No paper ever concludes by saying "yeah our method is perfect and no further work is needed" [1]. Instead, every solution has its quirks and questions which need to be explored further. Maybe their method has limitations which make it unusable for your applications, maybe they make somewhat faulty assumptions that don't always hold, maybe they wrongly ignore some technique. Seeing how other people approach a problem can often give you inspiration for how to take it another way.
[1] https://xkcd.com/2268/
I was surprised that writing the thesis was an immediate success. I've seen many PhD students struggle at this point, taking > 12mo to submit.
The 3 paper requirement in the game is also not a formal requirement in most universities--it's more of an implied requirement by individual PhD advisors. FWIW, my first lead-author paper I published a year past my PhD. During my PhD, I produced two relatively large scientific software applications (one open and one closed source) and a few open datasets. I'm now 8 years past my PhD and relatively successful in my field, 90th or so percentile based on common metrics--papers, citations, and funds raised.
Bottom line, papers are important but not the only thing that counts. Outside of tenure-track careers where they are crucial, it's possible to establish yourself as a scientist and be respected by your peers by publishing software and data.
Maybe an endless generation of MUD + LLM are actually the future of gaming