A New Way to Raise Test Scores
Last spring, my fourth grade daughter came home with her score on a district-required Reading STAAR practice test. (STAAR is the standardized test for Texas.) I was surprised to see she scored a 72, considering she is an avid reader and had reported to me that the test was easy. Was her score the result of gaps in my daughter’s reading skills? Were they the result of a poorly-written test? To my surprise, her score was the result of something else entirely.
I scheduled a conference with my daughter’s reading teacher, Ms. Horne (not her real name), to review the test. At the appointed time, the teacher and I sat on tiny chairs and dug into the questions my daughter got wrong. The first one was wrong because my daughter circled B in her test booklet and filled in B in the bubble sheet, when the answer was actually A. “OK,” I thought, “we may need to work on her inference skills,” as the question had required her to infer the motivation of the protagonist in the text.
We moved on to the next question my daughter got wrong. This question was a fairly simple question in which my daughter had to confirm a detail of the reading passage. My daughter was usually pretty good at these types of questions, as she loved to retell her Warrior Cats series plots to me—down to the very last detail.
It turns out my daughter had circled the correct answer, C, on the test booklet, but incorrectly marked B on the bubble sheet. Her error there was not a lack of skill, or even a poorly-written test. It was an error in filling out a bubble sheet.
Honestly, I was a bit relieved. But then we went through the next five questions my daughter got wrong, and they ALL were the result of improper transfer of answers to the bubble sheet. That’s right, my daughter circled the CORRECT answer in her test booklet, but filled in the wrong answer in her bubble sheet. Had she transferred her answers accurately, she would have scored in the 90s, not a 72, on her test! My relief turned into disbelief.
Do you know how many policy, student promotion, and even teacher compensation decisions are based on scores of tests exactly like the one my daughter took? It’s a lot. (See https://www.edweek.org/ew/issues/accountability/ for a synopsis.) It is hard to believe that our great centers of learning are largely based on how well a 10 year old can transfer a CORRECT answer from a test booklet to a bubble sheet.
Poor Ms. Horne may have gone to bed the night the test scores were released thinking she had not taught my daughter the requisite reading skills for fourth grade, when it was actually something as trivial as transferring answers to a bubble sheet.
I know standardized tests aren’t going away. But we could make them less of an anathema if the education system took the following steps:
Redesign the test so that machine scanners would read the test booklets where students typically circle the correct answer. This would obviate the need for a separate bubble sheet and remove a source of error, thereby making tests more valid. At the very least, do this at the elementary school level, where students new to such tests are probably more likely to make such mistakes due to their developmental levels and less developed motor and executive function skills.
Better identify the problem of incorrect transferring of answers by spot checking test booklets to see how many students, like my daughter, would have gotten a higher score had they transferred their answers correctly to the bubble sheet. Publish these findings so that parents and policymakers understand the impact of this issue.
The great developmental psychologist Jean Piaget once said: “The principle goal of education in the schools should be creating men and women who are capable of doing new things, not simply repeating what other generations have done”—and I would add— “or transferring answers correctly between a test booklet and a bubble sheet.”
