PARCC Opt-Outs Raise Question About Score Validity
As the opt-out movement continues to gather media coverage and spur debate, an interesting question is hovering about a particular group of students who didn't opt out.
That's right: I'm talking about the students who went ahead and took the PARCC tests. The question arises because of a quirk in the design of the those exams.
PARCC's English/language arts and math tests are taken in two distinct testing windows. In the first window, students take a performance based assessment. In the second window, they take an end-of-year exam. Together, the PBA and EOY make up the "summative" PARCC exam.
So here's what I got to wondering: If a student takes the performance-based portion, but then catches opt-out fever and skips the end-of-year part, does that student generate a valid test score?
One assessment expert says that, in essence, both parts of the test are necessary. Gregory Cizek, a professor at the University of North Carolina, checked with PARCC officials who told him students who take the PBA but skip the EOY test won't get a PARCC score. (Cizek serves on the technical advisory committee of the other testing consortium, Smarter Balanced.)
"From a technical perspective, enough 'information' about a student in terms of their responses to test questions is needed in order to report a score that is reliable. [T]he PBA and the EOY together provide a sufficient amount of information," Cizek told me in an email. "However, without one or the other, there isn't a sufficient amount of information, so the professionally responsible thing to do is to NOT report an unreliable score."
By the same token, students who have already opted out of the PBA have essentially ensured that they won't generate a score from PARCC, Cizek said. In any case, it seems unlikely that students who've opted out of the performance tasks would then decide to take the end-of-year.
How many students will end up with insufficient score information remains to be seen. Remember, the dates of the testing windows, as well as how much time there is between them, vary by state. In New Jersey, there's a 17-day-gap between the PBA testing window and the end-of-year window. In Illinois, meanwhile, there's only a three-day gap between the two windows.
PARCC hasn't responded to my request for comment about this issue.
PARCC is being given in 10 states plus the District of Columbia. If you're curious why I didn't discuss the Smarter Balanced exam, it's because Smarter Balanced only has one testing window.