Getting into a PhD program with a low GPA
3dpancakes.typepad.com
3dpancakes.typepad.com
So I took three of the hardest graduate CS classes I could find as a "non-matriculated student" (basically no degree, you just pay as you go for "professional" education), got a 4.0 that semester, and got each one of the three professors to write me a recommendation. Next semester, I was in grad school.
I maintained 4.0 for one more semester, until it slipped to 3.9. I always proudly tell this story every time someone talks about undergraduate GPA as a predictor of anything. (Actually, I would much rather hire people with a low undergrad GPA that shot up in grad school, than a high GPA all around).
I knew exactly what I wanted to do with my life (software), but I fell prey to the "college is required to succeed" dogma. I went to university for a year, spent all my time hacking and neglected the boring/redundant classes, then left after a year with a 1.7 GPA and took a coding job with a big salary.
Class was a waste of time.
College is such an easy life if you don't need to work to much to support yourself, many are happy just staying where they are and are fairly unconcerned about failing things.
I went to a liberal arts school, so there were a lot of classes I had to take that were completely unrelated to my major. These were the worst and the least interesting to me in general, but I had enough free time from not having to study for my major that the extra work required to get by in those classes (it takes a lot longer to study something you don't care about) wasn't a problem. I got 3 non-A classes: 1 for missing too many classes (always hated morning classes), 1 because the professor didn't understand basic math and nobody got an A, and 1 because I got the date of the final wrong, missing it by a week (yeah, that really hurt, my only C).
My girlfriend, OTOH, went to school for an engineering degree mostly because that was what her mom pushed her for; she didn't really know what she wanted to do. Her financial aid didn't cover room and board so she had to work nearly full time at Starbucks. She was a very hard worker and her managers loved her, but school often fell by the wayside, partly she didn't have enough free time to give it the attention it needed, and partly because there was a lot of family drama that was affecting her emotionally at the time. She now works at an engineering firm where they love her because she's a fast learner and incredibly hard working, but her GPA is low enough that she's not sure she could ever get into grad school.
The lectures who have taken the time to write their own lectures always have more interesting lectures worth going to.
Some I've had are extremely interested & active in their field, and they're all fascinating to talk with, and have useful, efficient classes. I've yet to see someone who isn't interested in their field do a good job teaching it.
If they are also pragmatic enough to see the utility of a degree, that would explain why they don't drop out of school.
I basically lived on my own schedule - I could be reading a totally unrelated book while everyone else studied for the finals. I made a point of being proud of it too, kind of silly in retrospect, but it worked for me at the time.
Going to class 3x per year: extremely low cost. Lost of free time. 2.6 GPA. Going to class every day: extremely high cost. Little free time. 3.x GPA (typically).
Given that semesters typically take 100+ days, you're talking a 30x increase in time spent to get a single point GPA improvement. Lame. This shows a problem with the classes and the teacher, as the teacher is apparently useless, not the student who is clearly smart enough to nearly BS a class and still get an easily passing grade.
Personally, I tend to fluctuate between 3.0 and 3.5. I do better in a lot of classes when I don't kill my brain listening to a bad teacher, and instead just read the darned book, which they test out of anyway. If going to the class is useless, and actively harmful, and you can get a good grade without going, what logic says you should go?
GPA is a laughable measurement of capability. It's much more of a measurement of boredom / lack of social skills. All my programming profs will (and have) happily write me a glowing recommendation, despite having a C average with one, and a low B with another. Of course, the C average one teaches level 450+ classes, and typically fails 50% of a class or more, so a C isn't too bad anyway.
I agree with your logic, although my experiences were different. There were definitely classes worth skipping, but I also had classes where it was worth it to attend. YMMV, I would recommend to anyone testing the waters before deciding on a course of action.
Oftentimes, it's not what you know, it's who you know.
I've always (always) had the stance that nobody can really teach anyone anything. All learning comes from the individual doing the learning, as it can easily be shut out if they simply ignore it. Thus, if you're learning, you're trying at least a little. (exceptions being made for "teaching" under torture)
People (teachers) and tools can make it significantly easier / better / faster than learning on your own from other sources, but ultimately you are the one doing the learning. The ball's in your park. I go to school / class because I see the value in it. I also slack where there is little to no value, or when I've learned all the class has to offer. To do otherwise is ludicrous, as I could be spending my time learning other things, which is significantly more valuable to me. My grade is temporary. Learning is not. Which is more important? Also, if an employer doesn't see the value in intelligence / learning, and only looks at my grades, I do not want to work there. They clearly have backwards goals compared to mine.
Grades, as pertaining to this, are not a measure of learning. They're a measure of how well you can match the test's metric, typically rote memorization of facts. Learning is connecting facts, and being capable of extending them, and that's hard to test. It also rarely requires real memorization, as you instead build off X to get to Y; you don't need to know Y, just the method, and the method could be applied in a thousand places instead of just the one Y.
If I can't demonstrate why I would be an asset, I've failed far more severely than a low GPA would imply.
"afford" is a pretty vague word. It depends on your priorities. Money is pretty low on my list. Happiness is near the top. I'll be more happy in a medium-bad job that I can easily change from, than stuck for longer in a medium job that I "had" to take. What cost are you measuring? Money? Free time? Ability to travel? Proximity to Wisconsin? (good cheese is important, dammit!)
Also, I'd like to point out that if they're not hinting to HR that they're interested in skills, not GPA, they're probably nearly a place I wouldn't want to work with. And / or HR is doing a crappy job, and they're allowing HR to continue to do so, which also puts them in that category. People are important, and any business that devalues people so far makes their employee satisfaction very questionable.
I am in CS because I like it. If I really need to, I'll pick something else up and do that. What matters to me is my life, my wife, and living happily, not making as much money as I can.
I took the programming job, and although it sucked, I got a lot of experience out of that job and am a better programmer today because of it. As horrible as it was, it probably sucked less than stocking shelves at Meijer would have, and I certainly got a lot more value out of it.
Last year we also had a group project which you could end up spending near unlimited time on, got a little unhealthy but I'd pretty much written up the whole core of it and was pretty happy with it so I kept pouring all my time into it. Also had an operating systems (mainly concurrency/compilers/memory management) which I loved and really aced. Also had a class on probability which was the first time I've had probability taken past the basic stuff you learn in general maths. Didn't get it well at first so it got next to none of my attention, didn't end up doing the homeworks which contributed to the end mark, crammed for it but still just failed. Hoping I find it more interesting this year.
Speaking from experience, probability and statistics will come up again in your life, and it is worth your time to learn them well. The answer to every practical question is "maybe," with varying degrees of certainty. You can't address these questions analytically without invoking probability and/or statistics.
I didn't take my statistics courses that seriously when I first took them, and have been making up for it ever since.
What is a 4.0 GPA in percentage?
A 95% average could be the same as an 80% or a 99% somewhere else. A 4.0 GPA could be a 3.1 or a 4.3 somewhere else (some schools use a "4.0" scale that goes above 4.0).
For example, I have a Masters degree, but I never finished an undergraduate degree (I dropped out). Similarly, though less bureaucratically, there are plenty of other dropouts working here at Google leading awesome projects, despite the 'must come from a top school or have a PhD' image it has.
It's cheesy, but in all such cases, I've seen that the biggest hurdle was getting over the "no you can't because you don't have X", either by trying harder, or by having someone else show you the way (or by dumb luck, in my case).
I've found that people tend to discount the past pretty heavily, at least in America. Nobody cares if you were lazy five years ago, as long as you've worked hard in those five years and have a record of concrete accomplishments to show that. If you went to a community college and then finished undergrad at Harvard, people will see the Harvard degree. If you went to a state college for college and then went to MIT for grad school, people will see the MIT degree.
The hard part is getting into Harvard or MIT or Google with that past record.
I think the important part for that is to not let your current circumstances define you. If you've been turned down for that dream job and have to work on CRUDscreen enterprise apps to pay the bills, it's really easy to sit back and get comfortable writing CRUDscreen enterprise apps. If you want to give that dream job another try in the future, though, it's important to continue working as if you got it, either by doing your current job to the standards of the dream job, or by keeping your skills sharp with side projects outside of work.
BTW, I think you were my phone-screener for Google, so maybe the article's point about having someone on the inside pulling for you has some merit.
For example, if you can get a job working at a university, that can make it a lot easier to get into a postgraduate course than if you're some stranger walking in off the street; and it can end up cheaper, or even funded through part-time work, etc. This is the de facto angle my girlfriend used to get into an economics PhD program at LSE, starting in a relatively low-level research position when getting out of the boring banking sector before the proverbial hit the fan.
But yes, I've heard of people doing this too and I can imagine it working. Professors who know you like you a lot more than professors who don't know you.
If you can get a research job and take class(es), you're practically a grad student already :)
As far as long PhD's are concerned, I think it's a difference in funding structure and expectation from some other countries. For one thing, when you hear that someone was doing their PhD for 7 years, that almost always encompasses master's work as well, even if they weren't awarded a master's. I am most familiar with how physics grad school works in the US, where the average PhD takes 6 years. Generally students who do get a master's on the way to a PhD get it after 2-2.5 years, so they spend 3.5-4 years on their PhD by itself. Another big difference from how things are done in the US versus what I know of in European universities is the way students are funded. Most European students I know are funded by the university/government and given 3-4 years to complete their PhDs. After that, they have no more money. In the US, funding is secured by the grad students themselves by essentially being hired by a professor as a researcher or working as a teaching assistant for a professor or the department (or possibly winning a fellowship from a government agency). Often this funding is on a semester to semester basis and in principle can last indefinitely. The fundamental nature of research means that it can last a long time when external constraints, such as funding, are not in place. I think my university had a rule that you had to graduate within 12 years or so.
I thought I'd mention that since the article (and others) mentions that it's harder to get in with a low GPA these days. This may be true, but it is still possible.
Your professors should know you after you finish taking classes with them. Preferably you should also be doing research with one (or hell, more) of them. Showing a professor your research potential is the quickest way to change your letter from "smart but lazy" to "smart and hardworking, but stubbornly doesn't care about the mundane class stuff that none of us really think is that important for research anyways."
As an aside: If you're interested in doing graduate work in security, please contact me. (My e-mail is listed in my profile.) We're always looking for good researchers and if you're good, we'd be interested in your application and it's possible I can poke someone on the admissions committee to make sure your app gets looked at thoroughly.
Also: Apply broadly. I had success at a few programs, some ranked higher than the one I'm in. I'm where I am because I followed funding and I wanted to be near the Bay Area.
And be persistent.
http://3dpancakes.typepad.com/ernie/2009/12/reading-phd-appl...
I dream of going overseas (USA) for my college studies, but due to the incompatibilities of our education system (Singapore follows the British education system), I don't think I'll get to use my diploma GPA for entry.
All grad schools care about is how well you did in that particular subject (or maybe a related subject, like math). You could easily ace your programming and math classes, bomb everything else, end up with a 2.0, and get into grad school. This isn't particularly surprising.
Wow, and you're proud of that?
About halfway through grad school I thought I might need to move for family reasons and thus had to apply to grad schools all over again. Having completed an MS by that point and a lot more research, very strong recs from people I had worked with extensively, etc., I got into every single program I applied to (including the top one in the country). I think my undergrad GPA was probably of little interest by that point, since they had a record of grad school performance. As it turned out I didn't need to move...