Hmm:
My success in any given class was almost wholly based on how well I could remember the definitions of countless terms
I find this somewhere between "unlikely" and "disingenuous", unless Penn State's CS department is just really shitty.
And then:
a 2006 study of college professors in STEM fields showed that a whopping 59.8 percent hadn't had any job experience in their industry. That means that a large portion of the professors tasked with teaching college grads how to become marketers, managers and salespeople have never marketed anything, managed anyone or sold anything at all.
Marketing, managing, and sales are "STEM fields"? Maybe STEM means "sales, technology, entrepreneurship, and marketing."
I don't get it either. If somebody wants to go to a trade school, it would probably be easier to go someplace else than to turn Penn State into one.
We are under some pressure to give students credit for completing a course if they can pass a standardized test somewhat related to the course produced by ETS. For instance, we were asked to start giving credit for having taken a general humanities course if you could pass a test on random facts associated with basic cultural literacy.
I actually took the test. It was all fact-recollection multiple choice questions, drawing on subjects from world history, dance, and architecture. I said there was no way we could say that passing this test was equivalent to taking our general humanities course, because the description of our humanities course says it will teach you writing, and this test had no writing component.
One place where this issue becomes serious is with people coming out of the military. Our Allied Health and Nursing program is supposed to give you credit for military medic training, but as I understand it, the training programs really aren't the same. The military program doesn't give you any foundations in anatomy and physiology.
All of this is just coming from people who don't understand the importance of basic skills, the kind that can't be measured by any one single test. I'm not really sure what kind of industries they work in, where these things aren't important. I'm also not sure why they think that all industries must work the way theirs does.
I'm not really sure what kind of industries they work in, where these things aren't important.
Marketing, managing, and sales.
Yeah, he could have saved me some time if he'd just condensed the entire essay down to "I wish I'd gone to business school."
5: Ah, yes, the STEM fields.
I blame silicon valley. They have everyone convinced that "working in technology" means "convincing a lot of people to buy an app that just says `yo'"
Without reading the link, which sounds stupid, it occurs to me that there's plausibly a case that college-level training in, roughly, "computers plus business" is hard to acquire. You can be a CS major if you want to be a developer, but actually most of the people at tech companies aren't developers, and many or most of them could really use a sophisticated understanding of the technology and a good grasp of how to e.g. do a sales visit or work business development. I have a couple of friends who make excellent money as sales engineers because they understand both of those aspects; neither of them went to college.
I blame fluoridated water. So many people are against it, it must have done something.
I guess you could just be a double major. That never occurs to me, because I'm lazy.
Do most bigger schools have some kind of design-your-own-major deal? Here at the University of Mn, it would be trivially easy to design yourself a CS/Marketing/Mangement/Sales major. I think you just need 2 profs to sign off on it, and you're good to go. Of course, I'm always going to be of the opinion that most kids who go straight to an undergrad bidness program would be better served by a 2 year basic office skills course and 2 years OTJ training.
Further to 4: I spend a fair bit of time with policymakers and practitioners in the workforce development world. My overall experience is that there is quite a lot of magical thinking going on, amplified by periodic panics.
When you feel responsible for helping a mass of people move into proper jobs -- whether you are a college administrator, local government official, or whatever -- you are going to look for the jobs that seem to have a clear link between What You Learn and What You Do.
You are going to look for jobs that seem to offer some kind of clear guarantee: I.e., if my students/clients get this piece of paper, they will be qualified for X kind of job.
The reason: There is virtually no other efficient way to get large numbers of people hired. And everyone, everyone, is worried about scale. You can get your friend's nephew hired through a one-to-one connection. You can't get all of your neighbors' kids hired that way.
By definition, this panicky focus will over-select for jobs that have very concrete, tangible tasks and established credentialing mechanisms. Does the world really need as many Certified Nurse Aides and HVAC technicians as we have? Maybe, maybe not -- but we'll keep sending people through training programs because for many people, the piece of paper is their only shot.
The big shift in the last 5 years, since the recession, is that this same kind of magical thinking is invading the college world. My organization is overflowing with bright young college kids begging to work for free* because they believe that they have to have on-the-job experience in order to get a job after graduation.
Employers, being no dummies, see this as a way to offload the costs of training entry-level white-collar workers. So now the expectations for a new hire are that you have ALREADY been trained/vetted via your numerous internships while still in college.
TL;DR:: If you think a job is dispensed by a magical Job-Creating Machine, you think students just need to push the right buttons. So skills like critical thinking, writing, etc. get shunted aside in favor of the flavor-of-the-month (social media expertise!).
*Free to us; they and their parents are paying $2,000-$10,000 to their colleges to get this credit
You can be a CS major if you want to be a developer, but actually most of the people at tech companies aren't developers, and many or most of them could really use a sophisticated understanding of the technology and a good grasp of how to e.g. do a sales visit or work business developmen
The weird thing (OK, one of the weird things) about the article is that he basically says he tried to get a job as a developer, but the languages used on the course were out of date. Which seems to be spectacularly missing the point.
Success in business and success in the classroom are certainly different things, but I'd wager that it's a lot easier to show an accomplished businessperson how to teach than it is to show a teacher how to be an accomplished businessperson.
Boy oh boy is this wrong.
Also, if you're going to write an article blaming your education for your difficulty finding a job, you should probably try to make sure that the reasoning in the article doesn't cause readers to think that the real reason people aren't hiring you is that you're bad at thinking.
Employers, being no dummies, see this as a way to offload the costs of training entry-level white-collar workers. So now the expectations for a new hire are that you have ALREADY been trained/vetted via your numerous internships while still in college
This has been the case in journalism for a long time, which is a large part of why it has serious class diversity issues.
if you're going to write an article blaming your education for your difficulty finding a job, you should probably try to make sure that the reasoning in the article doesn't cause readers to think that the real reason people aren't hiring you is that you're bad at thinking.
This, exactly.
I find his descriptions of what he learned to be a little unsettling. He learned how to explain "project management" in paragraph form, or the all-too-subtle differences between marketing and advertising but explicitly not how to program computers. I bet Penn State is pissed that this dude just admitted he got a CS degree without being able to write an app or program.
I believe that publishing mind bogglingly stupid ruminations about education has long been one of WaPo's core competencies. Are they still a Kaplan spin off?
As far as marketing and sales now being defined as STEM fields, I suppose once STEM became a magic word among administrators and policy makers, it was only a matter of time before the definition was diluted to the point of being meaningless.
19: I suppose that might explain why so many people with CS degrees were unable to pass the famous FizzBuzz test.
20.1: Nope. The Grahams kept the profitable businesses and sold only the Post to Jeff Bezos.
There's a Kaplan office in my building. Should I go pee on the door or something?
coming from people who don't understand the importance of basic skills, the kind that can't be measured by any one single test
I see that AcademicLurker got there first with the Kaplan connection. Does that persist post-Bezos?
Anyway, the answer to every suggestion for how to transform higher education is, when the parents at Princeton and Harvard think it's good enough for their kids, you can bring it to my school. Otherwise, this is all "let us teach you how to better gather your crumbs."
19: I suppose that might explain why so many people with CS degrees were unable to pass the famous FizzBuzz test.
Although I'd have thought Penn State grads would be pretty good at the Fizz Buzz drinking game.
He doesn't have a CS degree; he has a Management Information Systems degree, which is a stereotypical signal of a larval stage pointy haired boss.
I'm gonna bet he didn't actually get a CS degree, but something like this.
I had never looked at the FizzBuzz test before. Woo hoo, passed it using Excel!
He may have a point that they don't teach the hot new languages. It sucks for the teachers that they would have to learn new languages every 2-3 years or so, but it probably would help the students get jobs.
Marketing, managing, and sales are "STEM fields"?
I have very recently sat through several university information sessions. I had never heard the term "financial engineering" before, but I am now well acquainted with it. I suppose it's only a matter of time before Creative Writing departments rebrand themselves as Literary Engineering.
he has a Management Information Systems degree
And he's surprised he can't find a job?
31: I can't wait!
In the mean time, I'm pondering whether there's anything we do here that would benefit from using Excel ROW function. I can't think of anything.
I'm so happy I never have to sit through university information sessions.
32: not really. Any decent programmer should be able to pick up a new language pretty quickly, and the useful parts of most languages aren't the languages themselves but the libraries and frameworks you use in them. And those change too quickly to be a sensible target for a college curriculum.
Huh, so he said he graduated at the top of my class in computer science but actually majored in Techbro Studies? Still, Josh's link explicitly says majors can pick a concentration in mobile app development and software design, so it looks like he really screwed up. Oh well, now he's an entrepreneur. A founder, even.
35: You should learn SAS. That's the future.
32: the New Hotness languages aren't really necessary. There's a solid core of languages that have been popular since around the mid-2000s: Ruby, Javascript, Java, C#, Python, C. All of those have thriving job markerts. There might be some programs using more academic-focused languages (which aren't really obsolete, just niche) but I don't think that was his experience.
Techbro Studies
How long before that becomes an actual major?
In order to concern troll the thread, I do think there is a problem in that "practical" IT degrees like that in 28 have curricula formulated and taught by people who have left industry a long time ago or never been a part of it.
ACC's "Web Developer Certificate" will give you all the tools you need to make a great Geocities site.
Since there is no God but there surely is a Devil, 39 must be true.
I suppose it's only a matter of time before Creative Writing departments rebrand themselves as Literary Engineering.
Well, psychiatry's a medical discipline and politics is a science, so why not? It has a rather 1984 sound. Didn't Julia operate some sort of steampunk machine for producing low-quality pop lyrics automatically?
43: You just don't know pure elegance.
43: You just don't know pure elegance.
I do. I learned symbolic logic using only the Main Method.
"History being a branch of the biological sciences its ultimate expression must be mathematical." Colin McEvedy, introduction to the Penguin Atlas of Ancient History, or as it is now known, the Penguin Bootcamp of War and Civilization Hacks.
Unless it's along the lines of "my worthless MIS degree taught me how to program spreadsheets in vbscript", which is basically the point of those degrees.
Moby tasks me. He tasks me and I shall have him.
You should learn SAS. That's the future.
This made me guffaw to myself, because job listings for new positions in my unit list SAS proficiency as a preferred skill. When in reality, we'd love to get someone with solid basic Excel skills.
38: Well, he *does* work at a place full of "high-end web design experts"! And they write all their stuff in HTML5, so you know they're hot shit! (Saying you write HTML5 basically means "I have created a webpage sometime in the last 4 years".)
I thought those Info Systems degrees were "I'm not good enough at symbolic thinking and logic for a CS degree, but there are a whole hell of a lot of jobs that need someone with a lot of familiarity with the user-end of computers."
I don't get 49. There's whole bunches of SAS I never use.
51: We're having trouble keeping junior people because you need to be proficient in SAS. It's hard to find people who are and will work for what we pay.
You never use the parts of SAS that generate Wrath of Khan quotes? Those are the best parts.
Unless it's along the lines of "my worthless MIS degree taught me how to program spreadsheets in vbscript", which is basically the point of those degrees.
Which, again, it really shouldn't be.
54: I taught myself basic SAS when I was at a terrible temp job. They certainly paid me crap.
There's whole bunches of SAS I never use.
You could at least ask them to storm an embassy or something just to stop them feeling neglected.
SAS is certainly king in my industry.
And I was, adjusting for inflation, better paid when I was down there.
51: Did you actually look at their site? (Shorter: it's kind of a mess.) They have "very particular skills" including custom web design, SEO, and marketing. He will soon be a very wealthy man.
They forgot the e in "Plate 'O Web Design".
61: of course! That's how I found out they're high-end web design experts!
(Actually, I just looked at it on my phone. On a real browser: wow, they've hit just about every recent design cliche they could. I guess there's an element of "give the people what they want" but still.)
I thought those Info Systems degrees were "I'm not good enough at symbolic thinking and logic for a CS degree, but there are a whole hell of a lot of jobs that need someone with a lot of familiarity with the user-end of computers."
There are a lot jobs between "real developer" and "customer support": quality assurance, entry level system administration, software testing, database administration, code monkey, which should be in the reach of a well-trained student with an associate's degree.
Does their website show up as poorly for everyone else who isn't using IE8? I'm getting lots of white text on a very light background, completely unreadable.
which should be in the reach of a well-trained student with an associate's degree a bachelor's in the humanities.
It renders okay on Chrome for me (color-wise), but it's still pretty awful in general.
64: right. As in, that is validation of my impression that Info System degrees were worthwhile things for schools to be offering students.
65: Works fine for me in Chrome on OSX. Now, based on the copy I think I might be better off if it were unreadable:
Step 3: You enjoy an influx of new Harrisburg customers
And that's it. Sit back, relax - maybe grab a drink. You deserve it.
Good christ. That site is like the unholy union of medium.com and the <blink&rt; tag.
Also it looks like maybe you could use the assistance of some high-end web design experts yourself.
I have no good solution to this problem. I only want to say that I find it really depressing that a good, but not great, student who graduates from Barnard and isn't a good networker may have trouble getting a job as an administrative assistant--never mind some kind of career job.
Also, I've had some job situations where the people who were able to think for themselves got themselves into trouble.
69: they are, but there's usually a big lag between what they teach and where the job market is. The best programs usually have partnerships with local industry.
Are graduates of Barnard known to be notably better at administrative assistant jobs than graduates of other schools?
apo and Moby -- SAS more than STATA? I'd work for crap for a while to get a more interesting job. I have STATA and STATA books. Need to take a real statistics course, of course.
Somebody should really develop a statistical language called STATASS
a) I hadn't expected any sincere responses at all, but uniform (and I am afraid to say, funnier and crueler) mockery of the "writer," his alma mater, and the Post.
Witt is as always kind and informative though.
b) database administration ... Associate's degree If your organization has data to speak of, and if this is the way you hire dbas, then your organization is dying.
I would love to see the originally submitted draft of this editorial.
I do most of my analysis in Stata but all my data set management and building in SAS. Because Stata is more limited, it's easier to do the types of analysis that it does. But Stata without considerable statistical skill isn't something that I would think would fit you for many positions. If somebody is good with SAS, they can get a job without knowing much statistically. That's because SAS is useful for data management.
Twerk twerk twerk twerk twerk STATASS!
78: Yes, entirely. (NB: SAS world headquarters is ~20 minutes from where I'm sitting right this minute, so factor that into the equation, I guess.)
All the cool people are learning R. I still haven't figured out what that's all about.
77: No, not particularly. I'm just using it as an example because it was given as an example in an article I read. "You know and English/Art History major from Barnard who can't get a job.. maybe college isn't exactly the best way to get a job, if that's your only goal"
I think that a lot of the skills you learn in school make you good at writing cover letters and submitting applications, but the people hiring for administrative assistant positions (unless you know them through some kind of connection--and networking is a skill that isn't always cultivated in college) frequently want someone with experience rather than simply "bright and educated."
R has a reasonably clean design. It only works well if your data fits comfortably in RAM, but that's less of an issue than it once was.
"Have urgent enterprise-scale data management and analysis needs? Tap STATASS!"
81: An economist friend of mine wanted me to do some stuff for a data set they were building at one point. That's why I have it. He uses exclusively STATA.
"You know and English/Art History major from Barnard who can't get a job.. maybe college isn't exactly the best way to get a job, if that's your only goal"
If this is the kind of job you were going to go for, I wonder why you wouldn't have done that kind of work part time during school. It seems like it would be easier to get experience that way than with a fancy BA on your resume. I graduated college with retail, administrative assistant, and receptionist experience!
You can build and manage datasets in Stata. It's just much harder. Nobody who has to manage lots of them would use Stata and people who manage lots of data are the ones who are going to hire somebody to help them do it.
I think that a lot of the skills you learn in school make you good at writing cover letters and submitting applications
Recently I learned that I have a (mostly pro forma, there's no promotion yet) "two-year review" here and I have to submit a CV, teaching statement, research statement, and a portfolio of stuff including materials from both teaching and research. I feel like I'm applying for the job I already have, and it's annoying.
Every other time I've had to write a teaching statement I had the benefit of never having taught.
And I just had to have a meeting with our department's director of administration who tells me I should somehow find some tasks to delegate to an administrative assistant.
92: can they write you a teaching statement?
88: Some people get a work study job in the library or whatever. And they don't know what they want to do when they get out. So, they just apply for stuff. Not a great strategy, but a lot of places (read: Ivy League--that's just because it's my experience) career advice for humanities majors basically consists of: get a graduate degree unless you've got the grades and internships for management consulting or I-banking.
When my BF's Dad graduated from university in the 60's, his work experience was teaching swimming at the Y (a full-time job), the company that hired him to do sales and marketing put him through a year-long in-house training program. And his grades were pretty crappy, because he was working full-time while going to school. Nobody would do that today, because they don't expect people to stay with a company for a long time.
92: I assume you already don't get or return your own library books?
Is the two-year review the one that can get you bumped from assistant to (untenured) associate?
Delegating stuff is always quite hard. My current trainee/assistant is actually good/competent, and yet I still struggle, because time to teach someone else how to do X is almost always much longer than time it would take me to do X.
The pharmaceutical industry defaults to SAS for legacy reasons and because there are mountains of detailed documentation for it. R is becoming more common and FDA is okay with it, but if something goes wrong with SAS, you call SAS. If something goes wrong with R, well, who do you call?
I'm not much familiar with STATA, but it seems to be concentrated in economics. It barely registers in pharma research.
If something goes wrong with R, well, who do you call?
The graduate student who will be your boss in ten years.
If something goes wrong with R, well, who do you call?
The Letter Police!
99.2: Not really economics, but social sciences in general. We have lots of data that ends toward the social science end of medicine.
There really isn't any SAS or STATA around here that I can tell. Not sure why. My (stupid) statistics class used SPSS.
97: No, that's four year. I think they actually get outside letters for that one; this is just an internal thing.
SPSS isn't something I see often. I could still probably use it if I had to.
96: I don't need library books very often. When I do go to the library it's usually for something completely unrelated to my work, or just because I need a break from whatever I'm doing.
Stata is used a lot more in schools than in the private sector, is my impression.
If something goes wrong with R, well, who do you call?
stats.stackexchange.com?
108: I don't have a real good idea of what the private sector does because my schooling didn't cover marketing.
In my biostatistics class they told us to use either SAS or Stata. Everyone used Stata because it was not entirely command-line-based. Then it turned out that any future employer would probably only use SAS.
I've taken two biostats classes. One used SAS and one used Stata.
The one that used SAS was using it because Stata really doesn't offer anything to compare with Proc Mixed.
Oh! Maybe I can delegate all the times when some postdoc or senior faculty member emails me and asks me to download a PDF file from a journal and email it to them.
You could have an assistant sign up everybody you've ever collaborated with for ResearchGate so the spam emails they send you be true.
114: Yes. You forward those emails to the admin, copying the senior person. Eventually the senior person may learn to email the admin directly. This is a part of my life that improved when our new admin started last year. (Again, it kind of kills me that that's her job, because she's smarter than that. But so am I, and if she wasn't doing it, I would be.)
114, 116: That's a fair portion of my job.
116: The thing is, these people can download the PDF file themselves with basically zero effort, even if they're off-campus; they're just too lazy to either log in to a VPN or the library website to do it. I've been sending them the files with passive-aggressive notes about how they shouldn't need me to do it for them because VPNs exist.
And you think you can do passive-aggression better than somebody who has been support staff at a university for several years?
118: Sure, but educating senior people is depressing and effortful. An admin gives you a passive-aggressive no-energy out. Bonus, if you get on well with the admin, you can bitch about the senior people with them. It's a crazy reason to hire an admin, but if you have one sitting there, it makes your life better.
120: "Administration position available: The successful candidate will have extensive experience in SAS and bitching about senior level employees."
83, 85: there are signs that R is starting to have its lunch eaten by scipy and friends. Unless someone takes on the unenviable job of making R truly friendly to multicore I suspect this will happen.
122: It would be fantastic if most of the existing R packages get converted to python. The only reason I deal with R is because some of the packages are so convenient.
I'm still unhappy with scipy/matplotlib plotting. I feel like making it matlab-like runs counter to user-friendliness for those of us who haven't used matlab much.
122: I don't think that's going to happen, though it might. Python is considerably faster, but statisticians fell in love with R back when computers were slower, so it's not like they're going to fall out of love now. It wouldn't be that hard to make an R binding for numpy.
123, 124: agreed on both counts. I have some annoying scripts that mix low level code outputs with R and python scripts (and a bit of matlab) for their disparate strengths.
Still, for things that fit the workflow, iPython notebooks are pretty handy.
Honestly I'm not happy with anyone's plotting, but ggplot2 is nice for some things
122: I understand that, but the problem is that R is deeply single core, and the future is deeply multicore and distributed. I suspect first R will lose the non-statisticians, and then the statisticians will start to get jealous of features that other environments have. At that point it will probably be much easier to add packages to Python that make them feel at home rather than fix the architectural issues in R.
The community that cares about R deeply is a couple of orders of magnitude smaller than that of scipy, after all.
This is an observation, not a wish. I say this as someone who feels the numerical python approach is the wrong solution (more generally than stats) but that it's winning. At some point that becomes more important than the tech.
124: matlab-like runs counter to user-friendliness for those of us who moved over from Matlab, too.
This thread inspired me to install Julia, finally. That went fine, but it turns out that what I actually want (a Julia magic that I can invoke from a regular notebook) isn't working right now and doesn't really have anybody working on it.
essear have you messed with Julia at all?
129.last is rather a personal question, don't you think?
I never even heard scipy.
The new "I don't even own a TV."
131: I mean, Julia hasn't been around long and is full of holes, but seems like the kind of thing essear would get his hands on at least to experiment with.
You can be a CS major if you want to be a developer, but actually most of the people at tech companies aren't developers, and many or most of them could really use a sophisticated understanding of the technology and a good grasp of how to e.g. do a sales visit or work business development.
UNPOPULAR OPINION: neither how to be a developer nor how to do a sales visit is something that should be taught in a college.
Sort of on topic, I wrote an image viewer in Python using matplotlib for graphics and, because this was a quick and dirty implementation, the GUI, but it's really slow to respond to user input. Is this most likely the fault of (a) Python, (b) matplotlib, or (c) my shitty programming?
If I'm the only vote that counts a/f/a unpopularity that's not an unpopular opinion.
129: fwiw, Julia is a project I've played with a couple of times so far, and both times the conclusion was : wow, great idea, to bad I can't actually use it yet. Followed by, you know, someone who wasn't already putting in 80+ hour weeks could probably make a big difference on this project...sigh.
134 gets it right. Corrupting the purpose of college and putting the risk and financial burden of training on those least able to bear it.
137: yeah that was where I got very quickly... "huh. Could... I get this running? I should not be trying to do that right now."
I've heard that Julia incorporates a lot of Dylan's positive features, IYKWIM.
Julia sounded intriguing, but since I'm lazy, I tend to favor languages based on how much useful stuff written by other people can be found by googling. Julia doesn't do very well by that standard.
130: it has mostly been picked up by the sort of areas that used matlab a lot, so far. But the general model extends well to R/SAS/etc users, assuming anyone cares enough to write the packages well. There is starting to be that interest, afaics. Could easily be wrong, of course.
"Assume" makes a SAS of u and me.
You don't have to use the matlab syntax in matplotlib, essear; the other syntax is called OO in the docs and ?pylab? When importing. Also someone has ported ggplot, though I haven't played with it. Ditto RPy, which I should get working to avoid 126.1.
I feel I should like ipython notebooks but I want makefile dependency checking in them. Maybe that's happened while I wasn't looking? Added at a tea & scons party?
143: the relevant point isn't that assumption, it's the next sentence
I'm amused that a bit of googling terms related to my field along with "Julia" immediately brings up a blog post titled "I've seen the best minds of my generation destroyed by MatLab..."
135: all three in reverse order of seriousness? What kind of GUI? No tk or wx layer?
145: I think the relevant points should also include something to do with path dependence. Because of all the work that's gone on before (we're still working with projects that started in the last millennia), we can't really use anybody who isn't knowledgeable in SAS. This makes for very little incentive to invest the time in learning other packages. In my more lucid moments, I realize I'm not the whole world, but I'm fairly certain that a large portion of several industries is in the same situation.
Honestly I'm not happy with anyone's plotting
I'm generally pretty happy with Mathematica's plotting, or maybe I've just used it enough over the years to know how to get it to produce what I want.
But matplotlib seems needlessly perverse, like if I want a contour plot of a 2d function I have to first make some awkward 2d array of x, y values and fret about which is the row and which is the column and... ugh. I'm not sure if 144.1 means I'm missing something here, so I'll have to look into that.
I haven't tried Julia. A postdoc I was working with tried for a while to convince me to play with it and I kept saying I'd get around to it sooner or later, and then he left for a finance job. So: Julia enthusiasts make at least three times my salary, generalizing from n=1. You should all try it!
||
Academic bleg: I'm writing a CV and have no accomplishments. How bullshitty can I get without being super tacky? In particular, I have only one publication (from an REU and I'm frankly embarrassed about it). If I have a paper I'm planning to submit soon but I'm still waiting for my advisor to proofread and make comments, can I list that as "in preparation"? Or does that mean something more official? Also, I'm not supposed to put "work experience" on here, right? Like, nobody wants to hear about my job as a records clerk at the county clinic?
|>
Moderately bullshitty. Yes. Not always. Yes. Yes.
Good luck.
150: I think "in preparation" is fine; it wouldn't mislead anyone into thinking it's already accepted at a journal or anything.
I've seen a lot of grad school applications with irrelevant work experience listed and it seemed a little odd to me, but I don't think it really hurts, either. What context will you be using the CV in?
If the paper is nearly completely, print up a draft with a watermark saying "Not for Citation" or something like that.
And actually, records clerk is very probably something they do want to hear about. I just saw "record" and thought it was a music store.
I'm going to a conference in December and part of the application for travel support is a CV. My advisor is tight with the organizers so it probably doesn't matter but I don't want to look like a jackass.
On a real browser: wow, they've hit just about every recent design cliche they could.
It uses the bootstrap framework, which is modern web design cliche in a can. The point of bootstrap isn't high-end web design, rather to provide a framework for cranking out quick and dirty webistes that don't look like shit.... although I can see that they put some work into making it look like shit.
I only recently started using iPython Notebook instead of just command-line and it's pretty great. It makes a fair chunk of the convenience disparity with Mathematica go away, and it's still a lot faster.
Also, I don't know exactly what an REU is, but if people didn't list publications just because they were horrible papers, there would be a lot of CVs looking very empty.
148: that's entirely fair, and that's why there are still Fortran and even COBOL shops (though not in the same areas). I'm thinking more of where the new blood comes from. When a language becomes something you learn to support legacy projects, it has a different path. I think R (among others) runs this risk if only because it's architecture becomes limiting in modern computational environments.
Or think about it academically: is it a language you would recommend a new grad student learning for their own work? I don't believe we are at "no" for most disciplines, but I can see it coming...
For my own work, being able to spin something up on a half dozen clean instances and have them talk to each other is becoming more and more compelling, and I think that a lot of stuff will be going that way. He'll, I know people where that number jumps from 6 to several thousand, almost unthinkable outside of dedicated HPC facilities a few years ago, and now you could do that with AWS spot instances today if it economically made sense....
155: Yeah, definitely list the in-preparation paper. I think the records clerk experience wouldn't help with anything in this context, but it's fine to leave it there.
If you think of R as something that might be used in legacy projects, I think you're running really far ahead of where the main pack of users are chugging away.
And I agree with 160. There are uses for which the records clerk might help, but I don't think this is one. And I've now changed my mind twice, so feel free to ignore me.
You wouldn't rewrite the core of R to be parallel and distributed -- you would reimplement the key numerical operations in a parallel and distributed fashion. You don't need ggplot2 to be parallel -- you need + to be parallel. Now maybe no one will do that, but there's no technical obstacle to doing it.
Further to 150, your CV will be viewed in its proper context; I don't know if you're competing with more senior people for travel support or if it's only available to students, but it's generally given in preference to students anyway because senior people should have more travel funds. And in the student context, one published paper and another one in preparation isn't "no accomplishments."
Oh hey my REU paper has been cited 8 times, and 3 of those are by people I don't even know.
I know R! I could put that but it wouldn't take up very much room.
It takes up more room if you call it "R-stat." Up to five times as much room, depending on font.
I agree with essear, definitely list the paper. Also if you've given any seminar talks definitely list those, even if they're just for a small grad student seminar and aren't invited. Also, don't stress about it too much, grad student CVs just don't have much and you're not going to look out of place by not having much. Really the advisor line is all anyone's going to care about.
I hadn't read 164, but I also agree with that too!
150: it kills me that you started reading the comments here when you did and still ended up in academia. I can't help feeling like we failed you somehow.
156: I had a suspicion it was a bootstrap site. Twitter did a lot of good work with it, but it's definitely time for people to branch out.
On the original topic, people who got bullshit degrees shouldn't complain that their degree was bullshit. If he wanted to actually learn he should have had a real major, but I guess he didn't want to work hard or get grades that weren't A's, so bring in the world's tiniest violinist.
Also, I have an info sys degree, you insensitive clods.
It actually didn't teach me much, but its been enough to get me past HR trolls who want to see a degree in something computer-related.
it kills me that you started reading the comments here when you did and still ended up in academia
Yeah, me too. At least I'm marginally STEM and I didn't go to law school!
Thanks for the input, you guys.
Now I feel a little bad about 170. But not that bad, because as you say there's nothing wrong with getting a bullshit degree in and of itself, and as you say it can even be helpful in bullshit situations (like HR gatekeeping), the problem is then complaining afterwards about it being bullshit.
172: It occasionally strikes me how odd it is that in the time since I graduated from college law school has gone from a highly prized meal ticket to "At least I didn't go to law school!"
169.1: I don't think there's anything monolithic about academia. Most of the reasons extensively discussed in this blog to avoid academia apply mostly to the humanities and social sciences.
171.1: Sorry. Should we go back and explain the hard parts of the thread to you?
163: right, and my possibly incorrect impression is that the code base is a it of a pita to do this with. So someone(s) got to care and put in (a couple years?) work or it's not going to happen, at least not properly.
161: I didn't mean to say that's where I am now, just that I can imagine it happening over, say the next decade. One thing that is accelerating this is that I'm seeing more collaborative research code bases, and a lot of that seems to be happening int the Linux + Python configuration. For example I've seen OS distributions dedicated to a fairly narrow research area: download this, you get all the libraries and toolkit you'll need already set up, oh and data sets and code from a few papers are available here. That kind of thing can be pretty compelling, especially for getting new people up to speed. And if a big chunk of the best new stuff is coming out on a particular platform it's going to make you at least look at it.
Otoh, I could just be out to lunch on this, or ahead of the curve as you say. I often can't do my computational stuff directly in these environments, I just use them for glue or analysis.
the problem is then complaining afterwards about it being bullshit.
What if the complaint is that the system is bullshit, and that we are forced to do bullshit things to function in a bullshit system?
I can see a certain validity to the complaint of "dammit, I bought the wrong falvor of bullshit!"
L. Definitely list the paper, I'd probably put it as "in progress" (or in review if it actually is, but sounds like it isn't yet). That goes double if the content is at all related to the conf. You want to go to.
Submitting to the Journal of Universal Rejection is a good way to get a paper under review.
Depending on how done it is and what your advisor thinks, you could even post the draft of your paper to your webpage and call it a preprint.
I'd go with 181 over 180, if I were me.
L.'s lousy (I'm sure it is not lousy) REU paper has been cited far more often than my sole publication thus far, which has been cited once, by a friend.
For example I've seen OS distributions dedicated to a fairly narrow research area
Full OS-based distributions are old and busted. Containerized distributions are the new hotness (he said, linking to his personal, containerized iPython distribution.)
183: And more than any single one of my several publications as of the last time I checked.
So has anyone inferred I'm on vacation yet? I haven't posted this much here in years, hope it's not annoying to jump in today.
Since the NSF only allows you to list on your CVbiosketch preprints that have been submitted to a journal, what we need is something like the journal of universal rejection but where you can tell them how long you'd like the refereeing process to take.
It occasionally strikes me how odd it is that in the time since I graduated from college law school has gone from a highly prized meal ticket to "At least I didn't go to law school!"
The numerous completely moronic law professors who started blogging after 2002 or so probably didn't have a lot to do with this, but I hope had something to do with it.
Also, imagine being a youngish lawyer during precisely the period of the collapse of the profession's prestige, the moment everyone realized that there was an obvious oversupply of even "prestige" attorneys, and the cratering of its job market. But I shouldn't personally complain, I got in just early enough and coasted on prestige schooling enough to find relative stability, and of course it seems like literally every other profession except computer programmer, "developer," (no real idea what that is), petroleum engineer and financial engineer went through exactly the same thing at the same time.
OK, enough law talk, you guys can go back to speaking the quasi-comprehensible language of computer programs.
Just send it to The Southwestern Naturalist.
SAS is proprietary. You ought to at least back it for that reason.
L, if you're still following, in my field, a CV for a grad student might include brief (3 sentence) research summaries for relevant research experience. So, a section titled Research, then something like
Undergraduate Research Assistant, Somewhere U, 2009-2011
Advisor: Prof. Difficult
Responsible for generating new protocols for analysis of X.
Graduate Research Assistant, Elsewhere U, 2011-eternity
Advisor: Prof. Demanding
Designed and executed a series of meaningless tasks. Continued ongoing projects towards curing cancer.
Not sure whether it works in your field, and this often is removed once there are just papers and talks to cite, but I've seen it for students.
Hmm, Google Scholaring myself, my solitary paper has been cited twice. W00t! But ... one of my best friends from my undergrad period has been cited 1100 times.
I was thinking the other day about how crazy it is that so many of you guys are scientists or mathematicians who publish articles. "Hey, here we are at the absolute cutting edge of knowledge, and I've gone beyond that and moved it forward." Crazy. I mean compare to law professors, who really do just literally bullshit for the sake of bullshitting, or lawyers or consultants or whatever, who bullshit (arguably to maintain the rules necessary for a non-violent, orderly commercial society but whatevs) for money.
I was only looking in PubMed. Google scholar shows I've been cited very much more often. Thanks.
190: I think I've observed before that "Law Professor" is one profession whose prestige suffered considerably due to the advent of the blogosphere. I think my example was that Ann Althouse has replaced John Houseman's character from The Paper Chase as my image of a law professor.
Unfogged does seem to home to quite a few lawyers who graduated law school just in time to avoid the collapse.
Also, imagine being a youngish lawyer during precisely the period of the collapse of the profession's prestige, the moment everyone realized that there was an obvious oversupply of even "prestige" attorneys, and the cratering of its job market. But I shouldn't personally complain, I got in just early enough and coasted on prestige schooling enough to find relative stability, and of course it seems like literally every other profession except computer programmer, "developer," (no real idea what that is), petroleum engineer and financial engineer went through exactly the same thing at the same time.
It also seems like literally every other profession puts you in less student loan debt.
This isn't going to turn into "let's have a thread where we all post our citation counts!", is it? For one thing, the numbers aren't really comparable at all across field boundaries.
If I could somehow stretch the definition of my work to include dinosaurs, I bet that would help.
Hmm, Google Scholaring myself, my solitary paper has been cited twice.
Dang, I've only got one citation.
My crappy law review note has been cited several times in losing briefs by desperate, unsuccessful lawyers.
Clearly we should all start a co-citation ring.
199: Henceforth, all commenters must append their current H-index to their pseud.
I would guess that Cosma is the most highly-cited commenter here. Looks like he beats me by a pretty wide margin; he even has one single paper that beats my total citation count.
33: They've rebranded it as STEAM. Science, technology, engineering, ARTS, and math.
Unfogged does seem to home to quite a few lawyers who graduated law school just in time to avoid the collapse.
The later graduates can't afford Internet access.
Google Scholar claims my paper has been cited 13 times, but three of those are the same thing and one is a mistake.
So we remove the 1 and the 3 from 13 and conclude that ned's paper has never been cited at all.
Hey L., can you drop me a line?
Mysterious! Line dropped.
I thought I couldn't play this game, but then I remembered that one of my comments was cited in a book. Does that count?
It's not so easy to compare since essear hasn't set up a google scholar page.
124 - Have you tried using Bokeh? I know it comes with Anaconda's python distro, so it should integrate with ipython/scipy/numpy pretty nicely.
This thread made me realize that I have no idea what's happening with my citations these days, and don't care enough to check.
That's the second sign that I probably won't return to academia. The first (stronger) was earlier this year a department I like and admire invited me to apply for a TT opening and I declined because I want to finish what I'm doing now. I guess the zeroth was not taking TT when I had the chance in the first place.
It's funny when you haven't really internalized that you've made a decision, but you have.
Nobody has cited me for anything, ever.
It's not so easy to compare since essear hasn't set up a google scholar page.
Ooh, that's a perfect procrastinatey thing to do right now: it's there now. But the INSPIRE database is much better for my field; google scholar overestimates citations for some of my older papers (it double-counts a few things) and underestimates them a lot for recent papers.
85 cites to my one law journal publication according to google scholar, but many of them are duplicates or quadruplicates and a few were published before my article was, so I mainly conclude that google scholar is not not very reliable.
so I mainly conclude that google scholar is not not very reliable.
That seems unwarranted. Why not go with the simpler explanation that time travelers are citing your work?
Right, the double counting gets better if people actually get around to merging their preprints and published articles in their google scholar page, but that's always going to be enough of a hassle that most people won't do it. In math the AMS has a high quality database, but it's only published papers so the information is a few years out of date.
My one publication (I was a fifteen year old data-entry and unqualified SPSS monkey -- the professor I was working for made me third author) has been cited 38 times. I hope the fact that I lost half of the dataset we were comparing ours to didn't affect any of the results.
I don't even own a preprint. My problem is that there is a hugely prolific guy with the exact same name. I have to add coauthors to get to my papers.
My paper hasn't been cited yet, I don't think. Damn you all.
essear handily beats Cosma on h-index, by the way. But Brad DeLong beats essear handily.
Huh? At least going by Google Scholar, Cosma's h-index is higher by 3. (Although his since-2009 index is 1 lower.)
Apparently John Quiggin beats Brad DeLong handily, too.
I recently had a tweet retweeted 1100 times.
Oh, maybe I was looking at the five year. I take it back. Cosma is cooler than you.
Didn't RTFA as I'm an utterly shit hotel wifi connection, but we've stopped recruiting from PSU CS. With one exception, everyone that we've interviewed have been horrible critical thinkers and trained to use only a very specific and small set of tools. (At least with those from $local_prestige_university, when they say they only know $language_of_the_day, we can say "now do it in pseudo-$foo, which has the following relevant language features.") Lots of deer-in-headlights expressions. And that one exception was at PSU Erie, which is an entirely different program.
And I've got all y'all beat: zero citations. Zero publications, even. Unless you count some pending software patents, which is surely the most pathetic three words ever strung together.
229: I think it depends on the nature of the bot. Twitter's default apps and website don't make it possible to see all the activity but I know the exact numbers fluctuated a bit, and a lot of that was probably related to spam accounts being close. 1100 is not a large number, but two to three orders of magnitude larger than what usually happens when I tweet.
230: if you had managed to get citations without any publications, that would be impressive.
On IP: if the applications all go through, soon I'll be in danger of that count surpassing publications . Not sure how I feel about that. Actually, I am pretty sure it's "depressed". I'll console myself with the fact that a bunch of them involve hardware, at least.
650 cites on Google Scholar, but mostly stuff I did way back in the day, including software patents.
Of course I've only published one paper in the last ten years. (But they want me to do a few soon. Bah.)
On the subject of "teaching reviews," I can't say I teach at all, but our reviews are so disjoint from what we actually do that it's more or less surreal. "Pick from these goals which have nothing to do with your job, and require funding we won't give you."
Still older post # in the thread, I remember when SPSS was the hotness and I used to be pretty good at it. Now I barely remember.
To the OT, the problem with STEM hiring is that HR departments use keyword searches and hiring managers are encouraged to submit reqs with keywords. So yeah, somebody who knows Ruby (which usually means Rails) and so forth pops up and the rest are cast into the eternal fire, which is more or less bullshit, because in six months or a year there will be a new hotness. That's a major reason why referrals from existing employees work better: "Is this person smart? Is this person an asshat?"
And an aside to Halford: a "developer" is just a silverback code monkey. Really alpha developers are called "architects." Beards are suggested but not required.
No cites for my sole academic publication. Sigh. Hoping for another acknowledgement in a friend's novel sometime in the next year or so though.
Really alpha developers are called "architects."
Soon to be stepping aside for a generation of "systems ecologists."
163, 177: I strongly suspect that R is going to be re-implemented in much the way 163 suggests. Something like Renjin (re-implementing the R interpreter on the Java virtual machine to handle memory and parallelization issues), or Radford Neal's pqR. The key to getting something like this to take off would be to make sure that the existing packages carry over seamlessly. My impression, last time I looked was that Renjin wasn't quite there.
Notebooks in python are a bit intriguing, but I'm not really seeing great advantages over just writing R Markdown.
Friends-and-relations in industry at $large_tech_companies are about evenly divided between the ones who do everything in python, and the ones who prototype in R and then have software engineers translate into some form of C, Java, or Clojure for the industrial-strength implementations.
Oh, and what are these "citation statistics" of which you speak? Seriously, what are they?
236.last: Google scholar says that paper's been cited more than 150 times! They must have something interesting to say.
If something goes wrong with R, well, who do you call?
I can hardly wait to see apo and B/r/i/a/n Ri/p/l/e/y duke it out on R-help.
This correction to the article in the OP is priceless:
Correction: An earlier version of this story's headline misidentified what the author studied. It has been corrected.
The title (but not the url) now says he studied "computer science" rather than "engineering." Although per 27, it sounds like that's not actually true either.
As far as marketing and sales now being defined as STEM fields, I suppose once STEM became a magic word among administrators and policy makers, it was only a matter of time before the definition was diluted to the point of being meaningless.
Wasn't STEM always pretty meaningless, though?
Wasn't STEM always pretty meaningless, though?
Sure. And I think I'm on record here as despising the acronym. But to the extent it ever meant anything, it was something along the lines of "majors that people afraid of taking freshman calculus and/or chemistry don't pursue". Now apparently it doesn't even mean that.
241 is a great description. I wonder sometimes what we share with chemistry which makes that description the right one. I mean, every department could teach freshman classes that aren't easy, why is it chemistry specifically that actually does so?
Chemists are sadistic.
...Materialist.
Looking for a grant makes them opportunistic.
They're standing at the bench perpetrating a scan,
so a foundation with money can be their man.
Somebody is selling the secluded lot by that assholes dump trash on. The one by the beer store. I wonder if you can actually build a house there or not?
There's also a lot that has no trash. It's only 25 foot wide, which is common for that street, but the lot is currently the yard of the house one side. It looks like when they were both same property, the owner did things that would make it hard to split. Like putting a door, a bay window, and a Dish dish right up to the property line or over. If you wanted access to the backyard without walking through the house, you'd be luck to get a house with 20 foot interior width.
I mean, every department could teach freshman classes that aren't easy, why is it chemistry specifically that actually does so?
I think it's in part because long ago chemistry departments accepted, or perhaps even embraced, their designated role as "the department responsible for weeding out pre-meds".
There's probably a bunch of money in that.
Funny, I'd say it's freshman physics and math that determine the STEM selection process. Engineers don't usually have to take chemistry, and I think most biology majors are supposed to take a bit of physics. Freshman physics looks different that most high school physics if it's calculus-based. General chemistry shouldn't look much different.
The pre-med weed-out is normally organic chemistry, which is typically sophomore level, not freshman. The general reasoning on why it's hard is that it requires a different sort of problem-solving ability that's not really amenable to rote memorization (this is stupid, and I disagree, but that's what lots of profs and students seem to think). Most pre-meds are biology majors, not math or engineering, though, so lumping biology in as STEM is sort of funny, since it's an incredibly common major but very few undergrads want to be biologists post-undergrad.
244: Sad face.
248.last: I was babysitting my friend's kid the other night, and damn, their rented duplex is narrow! It's one of those house with a couple of later additions, both of which are narrower than the original portion, but even the original bit is only about 18 feet wide at the most. The housing stock in our neighborhood is weird -- lots of built-on-spec immigrant worker housing like that, and then here and there a block of really nice American Foursquare houses with all mod cons (ca. 1905).
Cob walls with insulation are going to need to be something like 2 1/2 foot thick. More if you want to use drywall on the inside.
Anyway, I'm going to change a ballast on the light in my stupidly not-cob house. But at least it's a house that is two rooms wide.
242: That article is interesting (though, indeed, pointlessly long and repetitive). What are the things it gets wrong?
251.last - Think of your profession as the sexxxy Wanda von Dunajew/Christian Grey of the sciences. Fifty Shades of Grignard! A thousand dreams that would awake me / Different colors made of transition metals!
Oh man. Yes, Harvard and Stanford undergrads are (practically by definition!) in fact the perfect suckers for the dumb status games of big name finance and big name consultancy. I AM HIGHLY INVESTED IN NAME RECOGNITION AND EXTERNAL VALIDATION! FIGHT FIERCELY!
257 is very true. I'd say there but for the grace &c., but in fact I'm not sure I didn't wind up on a different version of the same basic path.
There is (or was some time ago, and I suspect still is) a similar situation at certain brand-name schools of government and public policy, which (rightly) pissed the faculty there off even more. I don't know, maybe they found a fix.
Why is UC Davis playing football against Stanford? Is that why somebody left for a school that has a team whose main failings are moral?
261: Ah right, they were just making a move into "big time" football.
They lost by more than the number of states. Because patriotism.
261: the same reason shitty football programs always play out-of-conference football powers: because they got paid to serve as sacrificial lambs. Meanwhile, C.R.U. won its game in, wait, what? Ireland? I guess it's extra child-rapey there, right?
Child Rapists Beat Florida Men in Ireland.
Wait, 20 feet is narrow? (My house footprint is 20x30).
I feel like I should multiply that by my citation count (27? what? Those fools!) or something.
There was a time when there were fewer states than that.
Specifically, the 1880s (and earlier), if I'm counting right.
The match was slightly controversial in Ireland not because of the participants (I didn't even know who they were) but because it made the big stadium unavailable for a Gaelic football semi-final replay.