Raw Score vs "adjusted"

For level 2 I had a good score. I was +70 on everyting except two 5% topic areas were 51-70. I didn’t feel that I did great right after the exam and felt more or less like I do now, moderately confident, but I won’t be shocked if I fail. So I always wondered how the adjustment proces (if it exists) can potentially affect scores because I didn’t feel like my raw score was likely the score received in my results email.

Sooo having said that, how much do you think scores are adjusted after the review of exam results causes some questions to be thrown out or in some cases points to be given? Are we even sure there is an adjustment process during the grading?If yes, How significant do you think the impact is? Could the Raw MPS be something like 60% and when we see our score sheets we see some revised figure that’s higher because of thrown out questions or universial points?

TL:DR (up to 5% of final score) Yes I believe this also… actually if you heard the term raw score on this forum its likely from my post in level 2 forum. The idea is that each question is analyzed individually based on regressions and identified as fair/unfair based on forecasted vs actual aggregate results. Out of 60 questions maybe 2-3 questions are removed from the scoring (everyone gets full points) maybe another 2-3 questions, two choices are also counted as correct… another theory is that a few strategic “very easy” questions are given more weight to represent a baseline of core knowledge. Within the process the ethics adjustment comes into play. overall this means that a “raw score of 35/60 on PM” otherwise 58% can turn into 63% just because of 3 extra questions counted in your favor. This is why it takes CFAI months to finalize the scores. I never scored about 60% on a level 2 mock last year and passed with most sections above 70% even though it was very rare of me to get 5/6 on any particular tem set in practice. IF CFAI set a hard passing score of “70%” the passing rate would vary widely year/year because the mean normal distribution of 100,000 candidates would move greatly between 1-2 questions of difficulty… 1/60 is 2% it’s very telling that the passing rate has been so steady over the past few years