A SYSTEMIC VIEW OF TEST FAIRNESS
Warren W. Willingham1
Educational Testing Service
My theme--as implied by the title--is that the measurement profession needs to broaden its thinking about the nature of test fairness. Test fairness has always been associated in the public's mind as well as in professional literature with variations in validity. The nature of and reasons for differential validity have been a strong personal interest throughout my career, so test fairness seems a fitting topic for my contribution to the fine set of chapters in this volume.
Why are measures more or less valid for different purposes, in different situations, for different individuals? I think back on my earlier work on a variety of projects: to evaluate the accessibility of higher education, to decide which qualities and achievements of applicants warrant admission to a demanding educational program, to assess the academic capabilities of disabled students, to develop principles of good practice in the assessment of experiential learning, to determine which students should be placed on academic probation, or to characterize a student's success after 4 years of college. Each of those lines of work involved judgments of fairness. Few of the judgments were clearcut or entirely satisfactory.
A salient reason for the ambiguity and the unease, I am persuaded, is that fair measurement is a good deal more complex than the largely technical interpretations we tend to give it. Involvement with colleagues at ETS over the past sev-____________________