Teacher licensing tests are taking quite a hit this summer, at least in respect to how good of a job they do assessing teachers' reading knowledge. Widely leaked versions of an AACTE examination of state licensing tests landed the first blow, though NCATE--which commissioned the report--has yet to actually release it. Now former Massachusetts commissioner for teacher licensing, Dr. Sandra Stotsky, is out with Why American Students Do Not Learn to Read Very Well: The Unintended Consequences of Title II and Teacher Testing.

By examining practice questions for the most widely used licensing tests, Stotsky finds that prospective teachers can easily pass exams with little or no understanding of research-based reading instruction. On the Praxis 0014 general content knowledge test for regular elementary teachers, only two percent of the questions address phonemic awareness and phonics, and just one percent of the questions address vocabulary development. The stats are barely better for the Praxis 0020, "Introduction to the Teaching of Reading"--which is, bear in mind, the test ETS recommends for reading specialists--on which only four percent of the questions address phonics, phonemic awareness, or vocabulary.

A few states, perhaps frustrated with the poor quality of these exams, have developed their own content area licensure exams (with the help of the NES) to test prospective teachers' knowledge of reading instruction. Most notably, Massachusetts and California have developed reading specific exams in addition to the required general content area test. On the California RICA exam, Stotsky estimates that fully 45-50 percent of the test addresses research-based reading components.

While the report only looks at practice questions and online profiles of the tests to measure content, Stotsky provides a strong basis for someone to really dig deep into this issue.