Dear National Assessment of Educational Progress Board Members:
I write to request that you not approve the proposed replacement of NAEP’s assessment of reading comprehension, a change that could go into effect with the 2025 tests.
Like every researcher who is interested in improving the quality and fairness of American schools, I depend on NAEP (often called “the nation’s report card”) to gauge how well they are preparing our children to become prosperous, competent citizens. Nothing is more important to that preparation than their success in reading and understanding what they read.
As you know, the answer that NAEP currently reports is: not very well.
Along with many others, I’m especially disturbed by the inequalities that NAEP currently reveals in our children’s reading-comprehension abilities. For instance, the gap in scores between our white and Black 8th graders is almost a full standard deviation. This gap stems not chiefly from decoding ability but from comprehension ability. It exists largely because of a differential in relevant background knowledge between Black and white students. The proposed changes in the NAEP reading framework include deliberately offering needed background knowledge in preparatory material before the student reads the passage. This is well intended since it seeks to equalize the relevant knowledge differential between the test takers.
Yet the actual readings that these students will encounter in schoolbooks and websites, as well as in newspapers and the rest of the “real world,” do not normally offer such elaborate aids to comprehension. On the contrary, I am daily struck by how much is taken for granted in these sources.
It follows that this revamp of reading assessment would make NAEP’s tests less, not more, accurate and useful as sampling devices. By adding these special background clues, the tests fail to sample what they implicitly claim to be sampling. They will become less, not more, predictive of real-world reading-comprehension abilities. Moreover, since NAEP reports on groups, not individuals, the innovation does nothing for the self-esteem or social-emotional well-being of students.
While I admire the urge to be sensitive to the varied cultural backgrounds that students bring to school, that does not change the school’s duty to impart the knowledge required to master our common language. That, too, is a sociological imperative, especially in a multicultural nation. Moreover, we know that our elementary schools can achieve high literacy for all students, no matter their home cultures, because numerous schools are currently doing so, much to the delight of the parents of these children, and much to the benefit of those children’s futures.
In short, this well-meant proposal to replace the framework that governs NAEP’s reading assessment is not helpful and should be disapproved. It would not accurately report reading comprehension ability. It would not accurately expose the unfair gaps in reading between groups—gaps that we know how to close, that schools should be encouraged to close, and that their customers, the parents and guardians of these children, wish them to close to improve their life chances.
In fact, it is hard to imagine any positive result from this innovation, except possibly to make school improvement seem less urgent—something that no patriot or student well-wisher desires. As you know, the international PISA tests rank U.S. 15-year-olds at No. 25 in the world in their combined scores in reading, science, and math. We now have the 25th best school system in the world!
Of these combined scores in PISA, reading is the single subject most predictive of overall performance. That’s because the very languages of the classroom and of the schoolbook need to be understood by all the students in the class. For that to happen, all the students of the class need just the sort of preparatory background knowledge that NAEP’s earnest innovators wish to add to their test items. That preparatory knowledge is key to effective schooling. But the classroom, not the test, is the place to impart it daily and systematically.
So says current cognitive psychology. For more on reading and background knowledge, one could profitably turn to a book by the distinguished cognitive researcher Dan Willingham: Why Don’t Students Like School? It contains the immortal sentence: “A reading test is a knowledge test in disguise.”
E. D. Hirsch Jr.