After months of arduous studying, the relief felt by AP students in mid-May – finally able to close every Khan Academy tab, sit back during class time and regain some normality within their sleep schedules – is fleeting. The stress of studying for exams quickly gives way to the anxious wait for score releases.

Thus, within this protracted period, we may question, “What really goes on behind the scenes?”
Well, among nearly 1,300 graders nationwide – and 20,000 across all subjects – Cherry Hill High School East’s Mr. William Semus discussed his 17-year-long commitment to scoring AP Calculus exams with Eastside.

While most readers spend just a week in late June to score tests, Semus’ role as a question leader requires him to fly out to Kansas City a week earlier for an additional seven days of work.

“For the first week, for one of the six free response questions, I am responsible for finalizing the scoring guidelines after drafts are given to me, and then really fleshing out what exactly is going to earn points versus not earn points,” said Semus. “We go through hundreds of actual samples and look for certain things we think would be good to train readers [with]. You try to get one that barely makes the cut on both ends to sort of delineate that line for the readers to say, ‘This is the minimum we need to see,’ or ‘This is close, but it’s still not going to earn the point.’”

The majority of readers – approximately 450 on-site and 850 online for AP Calculus alone – then arrive during the second week to actually score the exams given the aforementioned new guidelines.

“We try not to be optimistic or pessimistic [and] play it as fairly as we can. Our general motto is that if a student has shown an understanding of Calculus, they should at least earn some points … We try to make the distinctions between a two and a three, or a three and a four clear so everything is set and graded the same way. Consistency is really key,” Semus said.

Creating exact guidelines for second-week readers is the most imperative factor in making sure the operation remains accurate and efficient. While these guidelines were limited to a singular sheet of paper prior to 2020-2021, question leaders like Semus are now free to be as thorough as possible, with the average sheet being four pages long.

“Where I think the advantage comes in [from a teaching standpoint] is that when a student might ask, ‘Would this earn credit or would it not earn credit?’ I have a better idea as to where that borderline answer would fall,” Semus said. “I also usually get a preview of things that are coming while I’m out there. For example, a year or two in advance, if they’re thinking about curriculum changes, we will often hear about those kinds of things first before that information becomes publicly available.”

These “borderline” answers are the main focus of discussion for graders and the driving force of the operation’s collaborative nature. Detailing “a tremendously collegial environment,” Semus described how each reader is assigned a partner with whom they decide upon point distribution. If uncertainty remains, the question is brought to a table leader and potentially all the way up to a chief reader.

“Although, yes, efficiency in terms of speed is important, the most important thing in our minds is getting the score right. So if it takes you ten minutes to grade a question, that’s fine as long as the right score is ultimately assigned,” Semus said.

Hence, as students cross the days off their calendar until July 7, we can find some comfort (or potentially discomfort) in knowing that each answer will be thoroughly examined and scored with the utmost precision.