>Universities should not be responsible for both education and assessment.

Easier said than done. In more niche disciplines, the only people qualified to evaluate student attainment are other academics in the same discipline. At UK universities, grades are routinely reviewed by an "external examiner" from a separate institution. In my experience, the result of this additional scrutiny has usually be to put grades up rather than lower them.

>> Universities should not be responsible for both education and assessment.

> Easier said than done. In more niche disciplines, the only people qualified to evaluate student attainment are other academics in the same discipline.

It is actually done in quite a few places. And even when there is no non-academic alternative, at the very least the examiners could entirely come from a different university than the lecturers.

In Germany there is the concept of "Staatsexamen" where a government-run examination agency recruits examiners from representatives of the profession and academia.

One of the many metrics that UK universities are assessed on is graduate employability - I think it's something like average salary 3 years after graduation in the latest Teaching Excellence Framework (TEF). Yes, we have a TEF to evaluate us, as well as external examiners and student surveys and other things.

Yes, I know. None of these things seems to have done much to combat grade inflation.

This is not a revolutionary idea. Several professional programs (law, accounting, etc.) have certification exams that do just what you are describing. It seems to require a fairly healthy, non-profit professional body in order for this to work, and you can only ever assess a fairly slim chunk of the 4-year experience, so they end up pairing in a credit-hour requirement as well.

I work in large-scale assessment, and can attest that we should hold a fairly skeptical/conservative position in the power of tests to get to the "real" skill level of prospective practitioners. All of the measures currently available are fairly weak at predicting outcomes.

Very much this. Removing the monopoly universities have on the (fairly questionable) credentials used for employment would go along way toward solving many of these issues. Most of the discussion about the problem with universities seem to dance around the core problem.

I doubt it's a coincidence that programs which require a third-party qualification exam tend to be more rigorous. E.g., engineering, nursing, medicine, law, etc.

The only thing I can say I learned from Engineering school is math. 4 years of my career completely changed my mind about what Engineering is.

I didn't know I would be solving problems.

Good that they teach math, but there's a giant disconnect between academia and industry.

In some ways, there is 'good disconnect' if the disconnect is that the academia merely informs your thinking, not your immediate actions. It shows that you learned the fundamentals in an insular environment that doesn't muddle them. If there is full disconnect, that is rough though

I Live in the Border city of Windsor to Detroit. We have a University that had been hailed for its engineering program. What I'm hearing from students/employers (big 3) is that the entire program is mostly mathematics, they have no usable engineering skills to begin work in the field because of minimal labs, and methods. Almost no CAD work/design control systems etc. They're so far behind the curve it's like they ignore that computers have automated so much of the math that they are teaching. More community college engineering graduates are getting jobs at the big 3 due to actually obtaining skills needed to do the job through labs and education that's on par with today's engineering.

> More community college engineering graduates are getting jobs at the big 3 due to actually obtaining skills needed to do the job through labs and education that's on par with today's engineering.

How does this work in Canada in regards to professional licensing? In the US, it my understanding that getting an engineering license requires an accredited program and that majority of a reddit programs are 4 year bachelors degrees. Even most master degrees dont have the right accreditation.

There are no US requirements for a PE stamp. There are 50 state level requirements, all somewhat different.

In my state, a four year BS degree in an engineering discipline is an automatic accept, although a non-engineering but engineering-related degree is accepted with four years of experience working essentially as a PE's apprentice. And regardless of degree situation, six years of PE apprentice type work qualifies. The criteria are not set by the state gov but by a professional cert board.

So a BS in EE from an ABET approved school is an automatic accept to start stamping civil engineering sewer projects, which seems a little weird as if obtaining a neuroscience degree qualifies one to perform open heart surgery, but whatever. My BS in computer science is "engineering like" enough that a four year apprenticeship under a PE would qualify me, although I didn't pursue that path. My SiL's primary K12 educator degree would qualify her after six years of apprenticeship under a PE, hilariously implying her liberal arts degree counts negative two years toward learning to be a PE, LOL.

I know some PEs and the required written test is reportedly very easy if you recently graduated, virtually impossible the longer you've been focused tightly in one industry. Stealth ageism much like how CS jobs "require" algorithm whiteboarding.

My state is extremely interested in prior felony convictions as a criteria. Not sure why that's relevant in any fashion to PE stamp a civil engineering project. I don't know how common that weird fixation is, this might be doc dropping what state I live in?

Anyway, in summary, there are 49 other states with differing opinions.

In Canada, you need a B.Eng if you're graduating today, or an equivalent from a foreign university+1 year local practice. There are engineers practicing in Canada that got grandfathered in on different academic systems, but there's not much leeway in the system these days.

There isn't the same disconnect in other STEM areas, a Biology or Chemistry student will do plenty of lab work to learn the techniques used in experiments.

That's a good thing, otherwise universities might be teaching 1950s "industry standards" that employers need to un-teach interns. Let the schools teach first principles and the industry shape new graduates.

The stuff you can learn on the job, you should. The stuff you won't is what college is for.

The weird thing is that people always seem to blame academia for the disconnect.

If an industry wants its employees to receive a specific kind of training, it should pay for it.

In my industry (infosec), they do and it works quite well. Fact of the matter is that you really don't need that much training for most jobs.