Both the President's State of the Union address and the Wisconsin Governor's State of the State address (and probably many other recent statements by public figures that I have not heard) called my attention to the problem of appropriate and effective assessment. If education is going to help achieve national and state goals, then the nation and the states will want some proof that it is doing so. Often acceptable "proof" involves scores on standardized tests, but there is considerable evidence in published papers in both chemical education and education in general that belies the validity of such "proof". Students work very effectively to find out what must be done to achieve with minimum effort the credential (grade, degree, etc.) that they want. The evidence is that they can do this on most of our assessments without genuine understandingthe kind that would lead to long-term retention and that would allow them to apply what they have learned to new circumstances. The common student complaint that a test question was not like any of the assigned homework problems, even when it was very closely related and understanding the concept involved would allow students to solve it easily, is one manifestation of this problem.

Studying for a test, and to a lesser extent teaching to a test, are long-standing traditions in secondary and higher education. These are not likely to go away, but we can use them to our advantage if we change what we say to students through our assessment tools. I am grateful to Edward H. Meyer, a retiree from high school and college teaching and before that from a job in the chemicals/plastics industry, for providing a thoughtful letter that directed me to a very useful book chapter on this subject by Senta Raizen (1), and also helped me coalesce some ideas that had been developing over several years. Raizen discusses the current movement to set national goals in education, including making U.S. students first in the world in science and mathematics achievement, and concludes that, "If the measures used to assess attainment of this goal are as narrow as most current tests, the effect on science education can only be deleterious."

Raizen sets some criteria that must be met for a test to be valid (2). The test should measure whether stated science education goals have been met. The test should mirror both the content and process of science. The test should require thinking and reasoning behavior that parallel the cognitive style of science. Use of the test should encourage and support good science education. By these criteria most science tests in current use, whether externally mandated or developed by a classroom teacher, fail miserably. Conceptual understanding and competent performance in science are not appropriately assessed and as a consequence are devalued in favor of memorization and rote "problem solving".

What can we do to address this situation? To begin with, a more comprehensive range of assessments is called for. Mixing several types of questions in a single test or exam, and using several types of assessment other than typical exams both contribute to broader, more authentic assessment. Multiple-choice questions are not necessarily bad, but they need to be made more complex to go beyond simple recall of information and address issues such as reasoning skills and conceptual understanding. The ACS DivCHED Exams Institute is moving in this direction with its new conceptual general chemistry examination. Essay questions can allow students to display their abilities to analyze situations, develop alternative approaches, and communicate effectively. Testing laboratory skills requires that performance tasks be developed so that students can carry them out in a real laboratory, not just write about them. One set of these was developed nearly a decade ago by a group of high school teachers under the auspices of the Institute for Chemical Education (3). Greater use of computers and information technology can be applied to assessment, tailoring tests to individual needs, presenting concretely situations that otherwise would not be available, and providing coaching and assistance. More details on each of these aspects of assessment are provided in reference 1.

Of course reforming the way we test is easy to argue for and a lot harder to achieve in practice. We have lots of experience doing what we have done before and lots of banks of multiple-choice questions that we can call on for tests of the type we have always used. How do you or I, busy enough with what we are already doing, change to a new paradigm that may require significantly more time and effort to implement? Probably none of us has the time single-handedly to attack this problem, but if we work together, we ought to be able to generate assessments that are a lot more effective than what we have now. One example is given in this issue, on page 528, and many others have appeared in previous issues. If you have a contribution to this effort, we certainly are interested in hearing it.

Literature Cited

1. Raizen, S. A. "Assessment in Science Education" in The Prices of Secrecy: The Social, Intellectual, and Psychological Costs of Current Assessment Practice, Schwartz, J. L.; Viator, K. A., Eds.; Educational Technology Center, Harvard Graduate School of Education: Cambridge, MA, 1990.

2. Raizen, S. A., et al. Assessment in Elementary School Science Education. A report of the National Center for Improving Science Education; The NETWORK, Inc.: Andover, MA, 1989.

3. Gardner, M., et al. Laboratory Assessment Builds Success, Institute for Chemical Education, University of Wisconsin: Madison, WI, 1990.