At the final PPL exams, about 80% dropped out, because they could not pass the CASA "trick question" examination.
Whether we like it or not and whether the comment is correct/fair or not, the reality is that there are two components to getting a tick in any of the CASA examinations (and I've sat quite a few over the years). First, one should know the material as that is the underlying reason for the whole exercise and, second, each exam is an area of study somewhat separate to the theory content and getting that elusive pass is a bit of an art form in itself. It doesn't seem to matter what area the exam is in or in what year we are looking ...
A recent area of interest to me is maintenance licencing and, in particular at present, the weight control examinations. They, certainly, are not at all technically difficult but the anecdotal suggestion is that the first time fail rate (especially in the second exam) is horrendous. Graduate engineers, who should be able to walk the work in, fall by the wayside in droves and the poor old maintainers also have a hard time of it.
Thinking back many years to when I was actively involved in pilot theory training, it was the same. Indeed, I did my pilot subjects 50-odd years ago and it was no different to now ... you learned the work and then learned how to pass the exam. Two quite distinct areas of endeavour.
Presuming that most folk who pony up to the exams actually want to pass, I suspect that the principal reason for the failure rate is that folk self-studying get the knowledge bit covered to a reasonable degree but then hit the exam without having done an adequate preparation in specific exam technique. For those folk who have done classroom study and still experience high failure rates, perhaps the theory instructors and training organisations are not quite up to scratch ?