Stakes don’t change scores

“High-stakes” tests are reliable measures of academic profiency, according to a Manhattan Institute study published in Teachers College Record. According to the study, the scores on high-stakes exams — tests that have consequences for students or their schools — track scores on “nationally respected tests that have no consequences tied to the results.” The study also concluded that “teaching to the test” helps students learn academic skills.

About Joanne


  1. John Doe says:

    High stakes or low, there is some randomness in scores on any given test, so scores on multiple equivalent tests are likely to be better indicators of mastery than scores from a single test.

  2. In Slovakia, like many commie countries, the final “test” consists of 3 difficult oral questions, from about 150. You know the questions at the beginning of the year. By the end, you’re supposed to know ALL the answers. Each students gets 10-20 minutes to tell the teacher their answer to the 3 randomly chosent questions.

    Not exactly fair, some Qs are easier, but teaching “to the test” means teaching the subject. I think this aspect is pretty good — there should be a fairly large, fairly standard set of “knowledge” that somebody who takes a class should know when they finish.


  1. What comes after you roll up your sleeves

    Via Reformk12, this story comes from Camille Cosby (Bill’s wife) about what it took to get her younger brother to read.