To: The Oregon Education Investment Board
The State Board of Education recently adopted a rule on teacher and administrator evaluation that can open the door to the use of standardized test scores as part of that evaluation. Such test scores would also play a role in other proposals, such as “common standardized formative assessments” and a longitudinal data base of individual student information.
Standardized test scores have built-in limitations. House Bill 2220, adopted in 2011, was cited by the Board in referring to a student’s progress toward becoming proficient in a “continuum of knowledge and skills” in mathematics and language arts. Test scores based on that continuum communicate nothing to a student’s parents other than whether the student passed, and they only communicate to teachers whether the student is near, above or below the acceptable point on that continuum for a given grade level. They do not tell what exactly a student knows or does not know, and if a student falls below the acceptable point, the scores do not tell why. In other words, the scores communicate no information on the actual standards the tests are supposed to represent
The “continuum” is a psychometric, mathematical model, with tests constructed separately for the supposed continuum of each subject matter. It is not a model of learning or knowledge.
Tests that do have diagnostic potential would be even more time-consuming than the tests schools are currently saddled with. Such tests would not tell a teacher anything the teacher cannot already know through the daily act of teaching. Test scores also do not tell a principal anything about a teacher that the principal cannot see through observation.
Moreover, standardized tests contain a degree of inaccuracy, called “error,” particularly at the student or classroom level. They are not sensitive to short-term growth. Again, tests that might be sensitive would consume even more time and resources.
The membership of the Oregon Education Investment Board is not selected on the basis of its knowledge of teaching and learning, much less the theory and construction of standardized tests. Therefore, as the OEIB considers proposals that have implications for more testing and use of test scores, I encourage the Board to seek the counsel of people who do understand the theory and limitations of standardized tests, in particular, people who do not have a personal or business interest in the decisions to be made. That would exclude test marketers and developers and those who sell the technology that testing requires. People who have that expertise can be found within Oregon public education and outside of the businesses that stand to profit from more testing.
Gary R. Hargett, Ph.D.
July 9, 2012
I am a self-employed consultant, specializing in the evaluation of programs for English learners and migrant students. I have formal training in test theory and development. I participated in the early stages of Oregon’s standard setting on the state assessments. I was a member of the original team that developed Oregon’s English Language Proficiency Assessment.