Is 40/60/80 a good measure?

I would say let us use a better one this time. If you get less than 50% … it implies 3 probabilities, you got 0,1 or 2 right on the vignette. Since we are at Level III it is hard to believe that all questions in the vignette will be wrong. Thus, either 1 or 2 correct. For 50-70%, 3 or 4 correct. And for >70%, 5 or 6 correct. To keep it simple if I go with mean in the three cases above, then it boils down to following weights: <50: 25% 50-70: 58.33% let us round it to 60% >70: 91.67% let us round this one to 90% So 25/60/90 for PM. Now AM, has more probabilities for each of the three categories. I would deviate again from 40/60/80 logic there but just for the first category (<50%). 40% is too high if you got less than half of a question correct. I would say let us come up with a better number here. Intuitively I will say it should be 30%. Thus we have for AM: 30/60/80. Any thoughts. Hope discussing this is not any kind of violation. This is what monopoly does - restrict the buyers. Pay for everything (exam registration, studying hours) and then do not discuss anything. I am fine till that point but then the test takers next year (hope none of us is one of them) will get this exam for practice along with the complete solutions. I don´t know what these kind of exams try to achieve by not allowing discussion.

am section: min random score 2005: 1 out of 3 questions : 14 1/2 : 15 12.16667 points by pure chance 2006: 1/3 : 13 1/2 : 16 12.333 points 2007: 1/3: 11 1/2: 1 1/4: 3 4.91 points 2008: 1/3: 4 1/5: 3 1/2: 9 1/4: 2 6.9333 points 2009: 1/3: 15 1/2: 3 1/4: 2 7 points On avg. over the last 5 years one could get 8.77 points by random guessing (based on multiple choice q-s within am section templates). If 1 point costs 1 minute then we get 4.87% ~5% correct by picking randomly. This way we get approx. 5%+ (50-5)/2 == 27.5% 28/60/80