In post-secondary education, there is a widely-held belief in a "gold standard" for evaluative studies of curricular innovations. In this context, "appropriate" assessment is understood to refer to experimental designs and statistically significant differences in group outcomes. Yet in our evaluative study of a medical undergraduate program, we did not find these concepts to be particularly applicable. Based on our experience, we now feel that it is appropriate to assemble an eclectic mix of scientific findings, show how they have been used for program improvement, and articulate the program's theoretical rationale and social significance. In the absence of statistically significant differences, this comprehensive argument can be used to justify the deployment of curricular innovations. The same may be true of other educational programs that target hard-to-measure changes in affective domains.