In the No Child Left Behind Act and the What Works Clearinghouse, the federal government has attempted to establish guidelines for the type of education research that U.S. schools should consider in selecting instructional programs and resources. The government's clear preference for the medical model--a powerful research design in such fields as medicine and agriculture--may be a mistake, writes Kingsbury. The conditions that are required for the medical model to work well include random assignment of individuals to treatment conditions; standardized treatment for individuals in each condition; use of a placebo that is administered to individuals in the control group; "blind" administration of the treatment and placebo; and standardized measurement of the outcomes. These conditions may not be obtainable or even desirable in schools. Therefore, educators should consider the information from a wide variety of nonmedical approaches in addition to the information gathered from medical-design studies.