CS Thread - How can one REALLY measure the quality between schools?

Considering last year’s average starting salary for RPI CS grads was $94K, UT-A grads probably make the same or less. My D is at RPI (engineering) but had best friends who are in CS. As with any university, you get out what you put into it, but for smart hardworking students RPI has very rigorous courses and a very high hire rate out of college. I’ve read a lot of comments online of transfers into RPI being surprised and how tough it is, and knowing a lot more once in their first job than peers from other schools.

Also bear in mind that different CS programs have different strengths. RPI has a strong hacker culture - thanks to a very nerdy student body - and is strong in infosec. For example, it recently came in #2 at the annualy Infosec competition at NYU (CMU first as usual), and also got picked as a sponsored team in the Amazon Alexa contest (again with CMU). It also has one of the best gave development programs (again, not a surprise given the nerds). But it might not be as strong in some other areas - for example UWash benefits greatly from its proximity to Microsoft, and Stanford to SV.

Adding to and adjusting @Jpgranier 's comment:

I’d say:

Group 1: CMU, MIT, Stanford, CalTech, UWash, Cal
Group 2: UIUC, UT-A, GTech, Michigan, UCSD, RPI, Mudd, and others

I think most people here are mainly in agreement, just with different ways of saying the same thing.

I think most kids do best with (A) a learning environment that agrees with them, and (B) peers who are about at their level. So of course it matters where you go to school - there are probably lots of schools that a kid would do well at, but likely lots of other schools which would be terrible fits. And so it’s not as simple as “go to the school with the best perceived reputation”. For one thing, reputation matters, but in CS probably not as much as people think it does. And it should go without saying that if you go to a decent school where you are outstanding and develop a love of the field and great relationships with a professor or two, you have come out way ahead compared to going to an MIT, struggling, graduating with C’s and no confidence / a developed dislike for your field.

I think what I say above is obviously true. Unfortunately it’s kindof begging the question. Because really what we’re all trying to decide is which school is best for our children, and / or is it worth the extra money. And quite frankly it’s probably impossible to know that in advance.

Personally, I’m much more interested in which companies come to their career fairs, as opposed to how they rank in (some listing), in particular because most of those listings are based primarily on graduate work, as opposed to undergrad education.

Has it been explicitly stated on this thread that USNWR does not provide a ranking for undergraduate computer science programs?

RPI, by the way, maintains a strong presence compared to all schools nationally:

http://www.businessinsider.com/the-50-smartest-colleges-in-america-2016-10

http://www.businessinsider.com/the-610-smartest-colleges-in-america-2015-9

I agree this is a pretty good indicator. Check the college websites as they sometimes have a list of companies attending their fair publicly available.

@thshadow @insanedreamer how do you go about finding which companies recruit where? I’m looking at a couple of my schools, but they don’t say individual companies.

If you’re looking at salary potential as an indicator then this list of desirable schools looks quite different: http://www.payscale.com/college-salary-report/best-schools-by-majors/computer-science

@Jpgranier I don’t know for other schools, but my D’s school had the list of companies attending the their fall fair on a page accessible to all. However, I just checked and that page is down now - and their spring fair list is accessible to students only. So I guess it varies.

@Jpgranier - just search for [ career fair]

e.g. rose hulman career fair shows from the first result:
http://www.rose-hulman.edu/media/1873284/2016-fall-cf-attendees.pdf

That’s a good example which demonstrates differences in opportunity. Rose-Hulman is a tiny school, so when you see how many companies go to their career fair, it’s obvious that there’s a very high company-to-student ratio – which is probably a good indicator of how companies value students in that school as potential employees.

@thshadow is right:

“Personally, I’m much more interested in which companies come to their career fairs, as opposed to how they rank in (some listing), in particular because most of those listings are based primarily on graduate work, as opposed to undergrad education.”

Getting this information should be part of your due diligence. It can be hard to get–often for good reason. @Jpgranier if the info isn’t accessible on the school’s web, you send an email to their career placement office. The reply you get will be telling.

I’m late, but just look up the “Annual Employment Report” for any college you apply to and it’ll show all employment information, employers, career fair recruiters, salaries, etc.

RPI’s annual employment report: https://www.rpi.edu/dept/cdc/Annual%20Report%202015.pdf

I read also many companies go to career fairs but ask the students to apply online. They usually receive thousands of applicants but will use a ‘bot’ to filter down to a manageable size of a couple hundreds for interview. I’m very ignorant in terms of current market behavior (being retired but a CS grad once upon a time).

How true is that for fresh grads or interns?

That’s pretty much the current CS market for fresh grads and interns too.
I asked Adobe a few weeks before for internship during career fair and while the person did not explicitly claim usage of ‘bots’, he did state the HR usually ‘filters’ resumes as they get around 3000 each year for internship for around 30 (in his sector).
The person seemed to avoid directly claiming usage of ‘bots’ but it was somewhat implied. He stated ‘they get the top applicants out of the resumes’ but it seemed an excuse to utilize bots to quickly remove resumes without checking most of them.

CS in bigger firms right out of undergrad for both internship and job seems more of ‘who you know’ rather than ‘how much you know’.
It’s quite saddening how many talented friends I know do not even get offered an interview because they have no connections over seemingly far less qualified students getting the internships and jobs simply cause of referrals, etc.

The software industry seems to be turning into (or maybe always was) connections over actual skill.
Either that or fill the resume with wasteful jargon to pass the filter bots.
It does not seem to value skills at all especially since fresh grads are all similar in terms of programming level.

Career fairs as a result of this in many universities seem to not help at all now. (also since almost all those companies ask to apply online instead)
Basically, to get a shot, one needs to either tailor his/her resumes with specific terms that would hopefully match with the bot filters (or it does not get read at all even if one is the most amazing programmer in the planet) OR know lots of people within the industry from the start which pretty much implies one already had enough resources before college to apply for such places.

And it doesn’t seem to help much frankly considering how interviews are done in unrealistic work conditions like writing perfect code on a whiteboard or over a very rudimentary site like Collabedit.
The current market really does not value much about skill as it does about connections or (bs-ing on the resumes) or having the ability to solve very elementary trivial programming problems ‘as fast as possible’.

Because of this, I really don’t think career placement states much about CS quality teaching in undergrad. But then again, in a macro perspective, it does seem better CS quality teaching schools should in the long run have far better career placements. So maybe it evens out. I don’t know.
Oh, and proximity to the companies help also if that matters.

I’m not seeing this. I’ll guess that less than 5% of the jobs I’ve seen filled in CS were due to connections. Most of the time you get a job by posting your resume on LinkedIn or a place like Indeed, or emailing a resume directly to a company or recruiter.

Seconding that I am seeing almost none of what @AccCreate is seeing. Yes, connections can help, but CS is still as close as you will find in any field to a meritocracy. None of my recent jobs have been through connections (the only ones that were my internships in high school) and I haven’t found any of my friends being filtered by bots at all. I know this because I actually know my resume is not easily searched by most bots (made in InDesign, not single column) and I regularly hear back from companies when I have applied with a good success rate. Adobe filtering resumes is not surprising, but there is nothing implying filtering by bots. I also know recruiters who are actually the ones who do the manual filtering at various companies, and it is just that - manual.

I’m sure there are some big companies that do bot filtering, but I don’t think it’s incredibly prevalent.

There are plenty of problems with CS hiring, especially when it comes to the technical interview and “culture fit” confirming previous culture biases and causing diversity problems, but resume filtering bots are not the big hurdle - even if used, it would be closer to chance than anything else given that resume design and CS skill are likely not correlated.

The only place I have ever witnessed referrals and connections making a difference is switching companies once you are securely in the industry - usually if a manager or director moves, they will take part of their team with them. Beyond that, at entry level, university recruiting is still the primary route, and if a company spends resources to recruit on a campus, they are going to spend some time with the resumes, not filter them through a bot.

What you are touching on is how hard it is to determine who is a good developer. Ask developers, and even they will struggle to identify exactly what it is.