INTRODUCTION

I was picking up my fourth-grade daughter after school one afternoon when, as she jumped into the car, she exclaimed, “Mom, I've never in my life had none of the above!” I thought for a moment and then realized what she was talking about – a multiple-choice test. Sure enough, as she continued to talk she expressed her frustration at a science test she had taken that afternoon. The teacher had decided to add the choice of “none of the above” to several of the questions, a choice my daughter had not understood. Never having seen it before on a test, she decided that it meant that she was not to circle any of the choices listed for the questions that offered “none of the above” as a response. Consequently, she failed the test.

This anecdote illustrates one of the problems found in contrived tests – including standardized tests as well as teacher-made tests such as the one my daughter took. In this case, the student knew the concept being tested, but was unfamiliar with the language and format of the test. Thus, her test-taking skills were what was lacking, not her scientific knowledge. Other problems that have been discussed in the literature with relation to traditional, standardized tests include norming on a population unlike the one being tested and cultural and language biases (García & Pearson, 1992, 1994; Wrigley & Guth, 1992).