ASSESSING AND CREDENTIALING LEARNING FROM PRIOR EXPERIENCE
University of Maryland
Some twenty years ago, Warren Willingham made a breakthrough early in the CAEL (Cooperative Assessment of Experiential Learning) Project ( 1974-1977) that was essential to its success and that had broader implications and applications that even today are far from fully realized. Reliability of assessment, he pointed out, does not require standardized tests and large numbers of examinees being compared on the same measures. It can be assured by way of processes that generate interrater agreement at appropriate levels. Validity of assessment, he added, does not allow examinees' knowledge to be judged on different grounds than their claims. He defined the crucial criteria for valid and reliable assessment in his Principles of Good Practice in Assessing Experiential Learning ( 1977). I assume that any sensible psychometrician already understood these ideas, but it was Warren Willingham who brought out the implications for measuring learning that had occurred outside of college precincts.
The implications and applications of the observation about reliability of assessment extend far beyond their use in assessing an entering college student's knowledge and other capabilities. The observation meant that it is feasible to provide reliable individualized assessment of learning derived from any source on a mass scale. I return to the importance of this Willingham insight later.
First, let us take note of Willingham's additional less heralded contributions and to some closely related matters that received less attention than they may have deserved. Sound assessment must be not only reliable, but also valid, Willingham reminded us. Validity being a matter of whether one is assessing what is intended, he pointed out that the strong preference of academics to judge students by what the instructors teach could mean invalid assessment if they rate the knowledge and competence claims of students in terms of what is covered in course syllabi and as-