No comment on the logic or the conclusion, but still, I couldn't resist myself form replying seeing the word "Statistics"Logic is flawed. And I'm sorry but saying that it's a fact, defies reasonable due diligence in making such claim.
One of my stats professor used to say "Statistics are like a bikini. They show what seems to matter, but hide the essential".
First, myimmitracker is not a reliable way to come up with the conclusion that September AORs are processed slower, as most applicants have not created a case on myimmitracker, and therefore cannot know the real number of September of AORs that got PPR or not by now. I'm not even mentioning the ones who have never updated their Immitracker ever since they've created them...
Second, your simple analysis completely ignores how accurate and complete September applications are, which may inherently delay the application processing (if there is any delay at all). Indeed, maybe the proportion of applications needing additional reviews is greater for September than for June, and maybe all applications from June were perfectly complete. That doesn't mean that the processing is slower...
Third, the fact that applicants from June for example got their PPR in 3 months, doesn't make it the norm, but on the contrary the exception. The normal processing time frame is 6 months, not 3 months. So, just because more people from June/July/August have it done in 3 months, doesn't make it a new reference for processing all of a sudden, and therefore claim that anything above that is from now on slow.
Maybe the proportion of applicants from September who get PPR within 5 months will be greater than the one from June, July, or August. You can't know until the 6 months period is reached for September AORs.
Look at it from a different perspective.
How about this one: June/July/August AOR was super fast!
Doesn't make September AOR slow. It's normal.
Firstly, myimmitracker is one of the reliable source of data. Statistical analysis include analysis of sample data which does not necessarily include or cover all data, and the sample size on myimmitracker is quite enough.
Secondly, analysis completely ignores how accurate and complete September applications are; True, but it also ignores such data of previous months AORs. How can you be sure that there are more incomplete / inaccurate applicants in Sept. when you does not the have any such data. Do not make poor assumptions.
Thirdly, Yes it is not norm that application should be processed in three month but that does but refrains anyone from making any claim, just like you made the claim "the proportion of applications needing additional reviews is greater for September than for June, and maybe all applications from June were perfectly complete"
We all gathered over here are in "waiting phase". Lets be supportive to each other.