The New Revolution in Testing

By Dr. Robert L. Brennan

Over forty years ago Sputnik was the impetus for great change, if not a revolution, in American education. Years from now, it seems likely that historians will declare that this too was a time of historic change in American education. What distinguishes the current revolution from previous ones, however, is the tremendous attention given to testing at all levels. This is most evident in K-12 with the passage of the No Child Left Behind (NCLB) Act of 2001, but there are also substantial changes in testing for college admissions and for professional licensure and certification.

NCLB is the newly revised version of the Elementary and Secondary Education Act, the centerpiece of the federal government's pre-college education legislation first enacted in 1965. In NCLB, the role of testing as an instrument of educational policy and accountability is unprecedented in this country. Consider two of the Act's provisions.

First, all states must administer annual assessments in reading and mathematics to all students in grades 3-8, and once during grades 10-12, by the 2005-2006 school year. These tests must be aligned to each state's own academic standards, and they must meet accepted professional testing standards. Meeting this mandate likely will require more test development, and more testing, than has ever been undertaken in this country in a comparable time frame.

Second, NCLB makes each state responsible for defining for itself so-called "Adequate Yearly Progress" (AYP) that must be achieved in annual increments so that all students are proficient under the state's definition of "proficient" by the 2013-2014 academic year. Furthermore, between now and 2013-2014, AYP must be achieved by each of a number of subgroups for the standard to be met. There are numerous philosophical, educational, testing, and statistical issues raised by this provision. For example, is it possible for all students to achieve proficiency in reading and math? Also, since all tests are fallible to some extent, what does it mean to say that all students are proficient on fallible tests?

Clearly, under NCLB the definition of proficiency, and the tests used to measure it, will differ by state. Acknowledging this fact, the NCLB legislation requires that states participate in the 4th and 8th grade National Assessment of Educational Progress (NAEP) reading and mathematics assessments on a biennial basis. NAEP is a federal testing program that provides estimates of nationwide and state-level performance in various achievement level categories (including its own definition of proficient). By federal law, however, NAEP cannot provide scores to individual students. So, in effect, the NCLB legislation mandates the use of a testing program that does not report individual scores to monitor each state's progress to a standard that every child be proficient.

While NCLB will probably have a greater effect on testing than any other single initiative in the near future, college admissions testing is undergoing a significant change as well. As a result of pressure applied by the University of California system, within a couple of years both ACT and the College Board plan to make the direct assessment of writing a much more integral component of their testing programs. Writing will be an optional part of the ACT Assessment but a required test for the SAT. Initially, trained readers will grade writing assessments, but there is a great deal of theoretical and applied research being undertaken in computerized grading of essays.

In licensure and certification testing, computerized testing is not simply a topic for research. A number of licensure and certification agencies have already begun administering their testing programs via computer, and other agencies plan to do so soon. For example, the National Board of Medical Examiners employs computers to administer the tests used to license physicians, and the American Institute of Certified Public Accountants will soon do so for CPAs. Relative to traditional paper-and-pencil testing, computerized testing is very expensive, but many agencies are willing to bear these costs in order to achieve higher levels of security and/or increased flexibility for their testing programs.

Testing can be a positive contributor to making good decisions about students, but such an outcome requires that trained professionals pay careful attention to professional standards. Adherence to these standards could be strained by the current dramatic increase in the amount and variety of testing at just about all levels of education.

Professor Robert L. Brennan is the director of the Center for Advanced Studies in Measurement and Assessment and the E.F. Lindquist Chair in Measurement and Testing. He directed the Iowa Testing Programs from 1994 to 2002.