A discussion on ABET accreditation came up on a student-alumni mailing list for Olin College, my undergraduate alma mater. Since ABET has been a tremendous shaping force for nearly all engineering programs and the vast majority of engineers and engineering students aren't aware of what it is and what it means at all, I decided to do a little background writeup for the list so the discussion would be more informed. I've made it more externally-understandable and reposted it here below.

First, I think it's important to note that "ABET" and "accreditation" don't mean the same thing. Most colleges are accredited by a few different bodies. For instance, Olin is located in the northeastern US, so in addition to ABET, it is accredited by NEASC, a regional body that grants it "real college" status (eligible for federal financial aid, able to grant student visas, etc - for a student's "why should I care" perspective on NEASC and similar bodies in other regions, see Kim McCraw's old Frankly Speaking article). And ABET, while known for its accreditation process, is actually a federation of 29 engineering-related societies like ASME, ASEE, IEEE, and others - so it does a variety of other things. So ABET is just one of many groups that gives colleges some sort of rubber stamp, and it does things other than accreditation.

However, when most folks say "ABET," they are talking about either the process of getting the +1 from ABET, or the status of having that +1. This rubber stamp is applied to an entire engineering degree program, not to individual engineers - it's a way to certify a course of study as something that "produces real engineers." Now, there are some famous engineering programs (for instance, Stanford) that are not ABET accredited - so clearly it's not the only way to be considered a "real engineering program," just like getting a CS or Software Engineering degree is not the only way you can get a programming job. But it is generally accepted as a screen of quality - you don't have a made-up degree from a sham school, you are probably a reasonably competent novice professional.

There are some clear advantages to ABET accreditation. Among other things, it allows a school's graduates to take the FE and PE exams that will make them eligible for the title of "Professional Engineer." (This is for the USA; other countries have their own variants and requirements for licensed engineers.) This may or may not be helpful to an engineering graduate. Of course, if you're using engineering as a jumping-off platform to another field such as medicine, business, law, (or circus performance, acting, and world travel, as some Oliners have done) this title doesn't matter to you or your future teachers and employers. It does matter to most civil engineers and to some mechanical engineers, largely ones doing work for the government; the exam and certification was originally designed for engineers and surveyors, if that's any indication. But by and large, you can work as an engineer without that certification without much trouble - some subfields (like my own, electrical and computer engineering) disregard it completely.

Accreditation also makes it easier for a program's classes to count for transfer credits, because other schools go "ah, yes, that class was from a real engineering school." This mostly matters for transfer students and those studying abroad. As far as Olin students at the time of our accreditation 5 years ago were concerned, those were the big salient benefits; we'd prove we could (and weren't unable to or simply being contrary), we'd get the ability to be certified Professional Engineers (which few of us have actually gone for), and credits would be easier to transfer (though it's not impossible to transfer them without accreditation - it's just sometimes harder). Do these reasons still hold? Are there new ones now?

There are definitely benefits to getting accredited; there's also a cost. The ABET accreditation process is 18 months long, which is about par for the course for academic accreditation processes in general; this isn't a trivial thing. You can read about the process if you're curious; basically, programs are evaluated by a jury of their peers, which in this case means professors from other engineering schools.

Far more important and illuminating is the criteria for accreditation. If you do nothing else before talking about ABET, read this. It is only 24 pages long (and the first 3 are cover pages, so really 21) and describes what ABET accreditation actually means and what it requires a school to do. (For Olin, which has 3 degree programs, the relevant pages were the introduction on p. 4-7 and the bits on Electrical and Computer Engineering on p. 11, General Engineering on p. 11, and Mechanical Engineering on p. 16.) It's important to actually look at the document, because without it people tend to make assumptions. For instance, did you know...

 

  • ABET does not require letter grades; it just says (p5) "student performance must be evaluated." How? By whom? How often? ABET leaves these as open questions.
  • Schools do not force (for instance) MechEs to take "outdated classes like thermodynamics" (as some students have grumbled) "because of ABET." Right on page 6: "The curriculum requirements... do not prescribe specific courses."
  • Engineering faculty don't need to do research, have PhDs, or even have academic backgrounds at all. Also on page 6.

Finally, the recent history of ABET accreditation is fascinating. It hasn't "always been this way" - the ABET requirements and process we know today were actually adopted (after furious debate) in 1997. Search for "EC 2000" (it stand for "Engineering Criteria") if you want to learn more; IJEE published a study of the impact of the changes (pdf) EC 2000 really started getting implemented in 2001 (coincident with Olin's Partner Year) and represented a big change in how engineering programs were evaluated. It used to be that ABET looked at "teaching inputs," which loosely translates to "students must take classes ABC." With the advent of EC 2000, ABET looked instead at "learning outcomes," which - again, loosely translated - means "hey, since the point of students taking classes ABC was to learn XYZ, why not just require them to demonstrate that they've learned XYZ, and let them accomplish that any way they want?"

So that's some preface. As for what I personally think... around the time of Olin's first ABET accreditation, professor Gill Pratt (now on extended sabbatical at DARPA) wrote a Frankly Speaking article on ABET, and one sentence sums up my sentiments exactly.

It is not accreditation we should worry about, but the aversion to risk that worrying about accreditation generates. -- Gill Pratt

That's why it's so important to be informed; when you don't know precisely what's required, you tend to err on the side of safety - which tends to make you exceedingly boring. After all, it's safe to do what everyone else does; they've got to allow that, right?

Sure. But there's so much more you can do - think of the requirements as a jazz fake sheet to riff from. If accreditation could inspire schools to push the envelope instead of staying inside it, if the process could challenge the boundaries of what we think of as engineering and of engineering education -- that would be a marvel. It would also take a lot of extremely brave people at all levels of the equation -- students, faculty, administrators, and members of the evaluation team -- who'd be willing to be brave, perhaps at high cost to themselves, for an exceedingly long time. And it might fail. It might fail.

Let us always keep ourselves in places of enough abundance to be able to take risks like that.