Andrew Gillen, Daniel L. Bennett, Richard Vedder
Measure What They Learned or What They Can Do Under the current system, measures of inputs and processes are used as proxies for quality. The final set of alternatives involves fundamentally changing that. There are two main variations. The first involves measuring what students learn, the second involves measuring what student can do.
The idea of measuring what students learn is most commonly associated with what is called a quali- fications framework. Modeled on a quality assurance mechanism common in Western Europe, a qualifi- cations framework would seek to determine what students should learn (in each degree or in general), and then evaluate whether the students actually learned it, usually through the use of qualifying exams.169
The idea of measuring what college graduates can do is most commonly associated with the concept of licensing or certification. Under this scheme, exams would be used to assess the competence of graduates on things that they would be expected to know or be able to do. For instance, in order to become a certified accountant, one must past the CPA exam that attempts to measure one’s mastery of accounting practices.
Determining institutional eligibly under these alternative proposals is straightforward—only those programs or institutions that do a sufficient job in increasing student performance on these exams would have access to federal funding.
These two approaches have a big advantage over the current system. Instead of measuring inputs and processes, and assuming that they lead to desirable learning outcomes, the outcomes themselves are the measure of quality. The main drawback is that this requires appropriately defining what is to be learned or expected of graduates and devising assessments that can measure learning or skills along those dimen- sions. Thus, these reforms are contingent upon being able to design an appropriate content, certification, and/or licensing exams that are widely accepted as both: a) testing for the appropriate knowledge and skills; and b) accurately distinguishing those that are knowledgeable/skilled from those that are not.
This is currently impossible for most broad categories. As one skeptic put it, “Let us imagine that pro- fessors of engineering around the world have been called together by government regulators to produce a single test or curriculum that will encompass every field of study within the engineering discipline… It would be a bureaucratic nightmare to come up with a definition of the field, or fields, to develop a test for all students and grade them on that test.”170 But even for a broad field such as engineering, there is already such a test, the Fundamentals of Engineering exam. Administered by the National Council of Examiners for Engineering and Surveying, the exam is used by many state licensing boards.171
Moreover, the chance of developing a successful exam increases as the topic of study gets narrower. So while it is highly unlikely that a suitable exam for something like the bachelor’s in business administra- tion (B.B.A.) could be designed, it is conceivable that one for a subfield like accounting could be. In fact, there already is an exam, the Uniform Certified Public Accountant Examination, for accountants. The bar exam for lawyers is another well known and widely used certification exam, and the Chartered Financial Analyst certification is becoming more common for financial analysts. Such qualifying exams are in widespread use in Western Europe.
Essentially, the basic idea behind these two alternatives to accreditation is to develop a CPA or bar exam for other disciplines. Any number of organizations could devise these types of exams, from exist- ing test makers like the College Board and ACT, to field based organizations such as the American Eco- nomic Association, to foundations like Carnegie, Lumina, and Gates, or even quasi-independent government agencies like the National Science Foundation, not to mention completely new organizations started for the explicit purpose of designing such tests.
The performance of a program’s students on the exam(s) could then be used as a measure of the quality