First tip – don’t bother with practice tests. MeasureUp has no practice test for this exam and all the other providers have a dodgy reputation. I did try TestKing but their software didn’t work at all, which was disappointing. This makes it really hard to find out what type of questions the exam will have. I searched StackOverflow for answers and the only thing I found there was a hint that using the older book (547) wont really help that much (http://stackoverflow.com/questions/963998/studying-for-the-70-564-exam).

The exam focuses less on code and a lot more on decisions based on requirements. So for anyone with lots of experience in 3.5, the exam will be a lot easier. This is probably the same advice as in my post about the 70-536 exam, but just going through the syllabus is probably the best thing. You can find it on the Microsoft Learning website (70-564 Syllabus). You can pair that up with “Programming Microsoft ASP.NET 3.5” for better coverage, but you can skip the basic sections and focus more on the high level stuff.

Finally, in that SO link above, theres someone who suggests that using the old book is no good. I can confirm that he is correct. Using the old 547 book would just be a waste of time. The exam is very 3.5 centric. This is a reverse of the first exam (536) for which one can comfortably study from the old book.

I hope that this helps any future MCPD-ers with their study.

I am perfectly aware that the score is ultimately not important; whoever looks at your transcript only sees that you passed, and since the pass mark is 80% and above everyone knows that you have done well at the very least. But I just can’t resist a good mystery. The following is a dissection of my personal score on this particular exam – I make a number of assumptions that may or may not be incorrect. Please consider that before you take my word as gospel.

The logic

There are 50 questions which means that each question is worth 2%. This brings me to the following assumption: Each question is worth the same amount of points

Seeing as the pass mark is 80% then 700 is 80% of 875. My score (826) is 94.4% of 875. This is an impossible score as that would require that each question was not 2% exactly. Hence the 700 is the fixed point and 0% and 100% are relative to it.

There are 7 areas in this particular exam and seeing as how I didn’t get zero NOR full points in any section, that would mean that each section comprises of at least 2 questions. This furthermore means that I got at least 1 question wrong in each section, which means that my final score was no more than 43 out of 50 (86%). We can also conclude that I didn’t get less than 41 out of 50 (82%) since if I got 40 I would be left with a score of 700.

Now we are starting to see a clearer picture. A score of 826 on this exam should mean a score of between 82% and 86%. But if you then try to find the upper and lower limit for the score you get some absurd results.

So where does that leave us? Either my first assumption is incorrect or I am looking at this from a completely wrong angle. However the bar diagram and the enigmatic score you are left with once you’ve passed your exam doesn’t have to be a closed book. Microsoft may have made the score as hard to interpret as possible to prevent cheating. Or maybe its very simple and the questions are weighted. Maybe with a larger sample of scores it would be easier to interpret. For now I must be happy with this.