The OP's '[2] Inaccurate Generalization' rider definitely needed to be there ;-) UK universities grade in many different ways, often based on course and project work. I'd say that the LSE's exam-only option is in the minority. For example my degree came out of a combination of closed book exams, individual & group projects, and open book essays. However it is rare for there to be continual assessment of coursework throughout the whole degree. It's generally based on a set set of special case bits of work near the end of a term/year.
That said — in my experience — the course work that UK universities that's much less susceptible to cheating in general than some stuff I've seen from US universities. Work varies more year-on-year, and the kind of work involved are things like essays, projects, etc that are harder to copy in a sensible way. I don't think I did a single multiple choice test in my whole degree.
He's spot on about the difference in expectation on top grades though. I've had some… interesting… chats with US students in the past when they went from all "A"s at their US institution to 70-80% at UK ones. Not so much because their answers were wrong — but because they're answers weren't 'excellent'. It seemed to cause an enormous amount of stress for them!
Most of my engineering and math courses worked on a linear curve. The highest grade in the class was made into a 100% and the rest of the class was slid up by an equal number of percentage points.
Sometimes this happened for the entire class grade, other times it just happened with tests.
In my general experience, a 70% was a pretty good score. Most profs aimed to have tests hard enough that 80% was around the top anyone got.
All that said, having everything rely on final exams seems foolish. Computer Science is mostly taught through projects, I feel as though my upper level courses had exams more as a matter of "well I think we are supposed to have tests, so, umm, here we go" than anything else.
Not to mention finals lead to cramming, which lead to horrible retention of material. Small cumulative exams spaced throughout the quarter/semester lead to much better recollection of material overall.
Edit: As for cheating, well....
I only ever cheated in one course, and only then because the material was horribly unfair and the professor was terrible at teaching. Cheating in this case consisted of putting material that we were expected to rote memorize (!!) on to my calculator. (For reference, it was chemistry class, before that class I used to really like chemistry... :( )
Asides from that? I think everyone knows of the student who seomehow got through even a hard program (such as CS!) without really doing his or her fair share, but from what I noticed (although I am pretty oblivious) cheating wasn't all that common.
In regards to grades, they mattered far less once I was in my department. :) (Thankfully!)
Another key difference that I couldn't see in the article is that degrees in the UK are focused on depth (and last three years); while degrees in the US focus on breadth and last four years. This probably also relates to how students in the UK specialize a bit earlier (during high school / A levels) while students in the US leave that decision to their first year of college.
That said — in my experience — the course work that UK universities that's much less susceptible to cheating in general than some stuff I've seen from US universities. Work varies more year-on-year, and the kind of work involved are things like essays, projects, etc that are harder to copy in a sensible way. I don't think I did a single multiple choice test in my whole degree.
He's spot on about the difference in expectation on top grades though. I've had some… interesting… chats with US students in the past when they went from all "A"s at their US institution to 70-80% at UK ones. Not so much because their answers were wrong — but because they're answers weren't 'excellent'. It seemed to cause an enormous amount of stress for them!