The differentiated outcome hypothesis posits that the maximum effects of an educational program are realized along those dimensions emphasized in the program or in those central to the persons involved. The idea that educational programs employing unconventional means should be examined on outcomes that are consistent with those means is derived from this hypothesis. In addition to knowledge acquisition outcomes, thinking and problem-solving, attitude and value, and behavioral change outcomes are recommended. Studies are cited that utilized outcome measures from the above classes and successfully demonstrated the differentiated merits of the programs involved. Relevant studies in progress are also reviewed. (Author)