WATER UNDER THE BRIDGE A Response to Bingham , Heywood , and White AURELIO JOSE

ncreasingly, powerful computers and statistical packages now offer ~~ researchers a wide range of statistical analyses capable of handling large numbers of cases and variables. Along with this cornucopia of statistical possibilities, however, certain problems emerge. Foremost among these is the possibility that researchers will end up using statistical analyses with which they have little familiarity, such as with respect to underlying assumptions and functional limitations. The use of unfamiliar statistical packages and procedures may, in turn, lead to inappropriate statistical analyses and unwarranted conclusions. The resulting problems can be very serious indeed, especially when conclusions are relevant to public policies. A recent example of what we think is an all-too-common problem is the article, &dquo;Evaluating Schools and Teachers Based on Student Performance: Testing an Alternative Methodology,&dquo; by Bingham, Heywood, and White (1991), recently published in this journal. Although presented by Bingham et al. as a new method of determining the influence of teachers and schools on student achievement, the statistical methods actually employed by them result in conclusions that are almost certainly wrong and that could result in very bad policy decisions. Their analyses are inherently unable to provide reliable and valid results because of a failure to recognize the statistical limitations of the procedures they followed. Because it may not be immediately obvious to the entire readership of this journal how the particular