Archived
K-12 Testing
By Samuel J. Meisels, Ed.D
Experts and organizations concerned with academic assessment generally agree that standardized, group-administered tests should not be used with children younger than third grade. In the past decade, advocates of appropriate testing have had some success in replacing such whole-group tests with direct, one-on-one assessments.
Aside from the move to individualized administration and the advent of some tests based on an IRT statistical model, little else has changed. Preschool and kindergarten children are routinely given tests that are thinly disguised versions of standardized multiple-choice exams. Most federal and state-mandated tests used for our youngest children are simply a pictorial version of the traditional “fill in the bubble” multiple-choice test. For example, in a widely used vocabulary test, the administrator says a word, shows a child a set of four pictures, and asks the child to point to the one that best represents the word.
There are many problems with this approach. The first is that young children’s development is extremely labile; it is in a constant state of change. Until third grade, children’s skills—and their ability to demonstrate those skills—change rapidly. No one-time administration of a standardized test can adequately capture this change and evolution.
Second, these tests rarely reflect the lessons and skills children are really learning in preschool and kindergarten. They don’t address physical growth and motor skills, social-emotional development, the arts, appreciation for reading, early writing and math skills, and more. They tell us nothing about a child’s problem-solving process or preferred method of learning. They ignore vast differences in culture and previous opportunities for learning that shape the context in which the child experiences the test.
These and other drawbacks are serious concerns because testing is an increasingly high-stakes proposition. In the current age of educational accountability, test results are used to make crucial decisions about student retention and promotion; evaluation and rewards for teachers and administrators; and allocation of resources to schools. Research shows high-stakes testing can make a long-lasting impact on a child’s self-image, motivation, and achievement. For teachers, it presents an irresistible motivation to “teach to the test.”
A better alternative for young children is observational assessment firmly embedded in the classroom curriculum and conducted by teachers over time. Unfortunately, instead of growth in observational assessments, we see increasing evidence of inappropriate high-stakes testing. Instead of instruction-driven measurement, we see measurement-driven instruction.
The most glaring example of this is the National Reporting System, which calls for testing the half-million four- and five-year-olds in Head Start programs twice a year. The test in use omits much of what is actually taught in high-quality Head Start programs. Rather than eliminating failures in the Head Start program, this test may teach young children to view themselves as failures, simply because they see things differently from the way the test developers do.
Although we do have screening and diagnostic assessments that are useful with young children, currently there are no norm-referenced achievement tests that are sufficiently accurate to serve the high-stakes purposes they are being asked to perform. Rather, we are witnessing a continuing misuse of tests in the early childhood years as the testing culture of the older grades invades the lives of young children.
• Dr. Meisels is president of Erikson Institute, a graduate school in child development in Chicago. He is a leading authority on the assessment of young children and the impact of standardized testing.