How Do Teachers Know What Students Know?
In dusting the shelves of my home library recently, I came across a paperback published in 1961 that has uncanny relevance to the debate today about mass testing (The Schools, Anchor Books). Martin Mayer was a reporter and editor who spent 30 months visiting about 150 schools from as far east as Helsinki to as far west as San Francisco. In the process, he interviewed more than 1,500 teachers. His provocative and insightful comments are a reminder that so much of what we consider new today has in fact been around for a long while.
In particular, Mayer reminds us that determining learning depends in large part on the theory we believe in. "From the point of view of the laboratory psychologist, it is convenient to define learning as 'behavior.' The psychologist needs a measurable result. That the definition is essentially a fiction is clear from the large group of phenomena which must be classified as 'latent learning' - that is, learning which does not show itself in behavior." In other words, students often know more than they can express.
This truism fails to satisfy reformers who demand empirical evidence of learning. I can understand their position. Taxpayers are entitled to know if their dollars are being well spent. In the minds of the overwhelming majority, the only way to know for certain is the production of palpable outcomes. Claiming that not all learning can be demonstrated behaviorally is seen as an excuse. Readers of The Wall Street Journal's editorials and op-eds are familiar with this position.
But I wonder if students are being well served by this belief. B.F. Skinner liked to point to a cartoon showing a rat leaning against the puzzle- lever in his cage and telling another rat: "Boy, do we have this guy conditioned. Every time I press the bar down he drops a pellet in." In short, the design of a test restricts the range of possible results. Whenever I took a standardized test requiring a selected response, rather than a created response, I never felt that my score reflected my true ability because I was hemmed in by the construction of the item. Is that an excuse or an explanation?
Of course, instruments that rely heavily on created responses by students will be far more costly to design, far more expensive to assess and far more subjective than the tests currently in use. But I think they're worthwhile considering in light of what is on the line.