Can a national text measure college learning?

YES
Margaret Miller
Seek reliable comparisons

Colleges differ so widely from one another in this country, as is repeatedly and correctly pointed out by opponents of standardized testing, that no single test can capture "college learning."

But if the question is, "Should we have some standardized measures of the general intellectual skills of college graduates that would enable us to compare student performance across similar campuses?"-I say yes.

First, American colleges and universities suffer from the rampant individualism that threatens the balance between the private good and the public good. But we are engaged in a communal enterprise, the creation of this nation's future human capital. We should periodically check to find out whether we, collectively, are doing that job well.

Second, our system has become so diverse that the very meaning of a college degree is in question. When we give someone an associate's or baccalaureate degree, just what should we be certifying about them? To begin answering this question, we need to find out what range of skills we are warranting.

Finally, despite a rich array of program-based assessments, faculty now can't determine if they are doing a good job of educating their students. Compared with what? How about, compared with other institutions that educate students with similar entering characteristics?

Does this mean that we need to have a single national test? No. The legitimate variety among our institutions requires a variety of assessment strategies, as long as reliable comparisons among like institutions are possible. The Voluntary System of Assessment being developed by the National Association of State Universities and Land-Grant Colleges (NASULGC) and the American Association of State Colleges and Universities (AASCU) will identify existing measures that might be used by clusters of institutions to look at their performance-just the right way to go about this task.

Does this doom us to multiple-choice tests? No. Standardized performance-based measures now enable us to determine our students' general intellectual skills by asking for demonstrations of them. And we may even find a way to make scoring rubrics for portfolios reliable enough to use them for this purpose as well.

Does this mean that we have to test every student? No. Sampling also has developed to the point where a small group of students on each campus can be assessed to determine the overall level of their group. This also keeps costs down and makes the process nonpunitive-differentiating this strategy from No Child Left Behind, which seems about to fall of its own bureaucratic weight.

NO
Catharine Hoffman Beyer
It misses the obvious

Using a national test to measure college learning is like holding an expensive thimble under Niagara Falls, so we can assess the experience of being there. We might learn something from the few drops we collect, but we will miss what is most important about the experience.
Findings from the University of Washington's Study of Undergraduate Learning (UW SOUL), which tracked 304 entering students as long as possible between 1999 and 2003, reveal two significant aspects of college learning that national tests would miss.

First, standardized tests would miss the full-frontal, transformative experience that college represents for most students. Entering UW SOUL participants told us they hoped college learning would bring a deeper understanding of themselves, greater skill in social situations, a post-college direction connected to their values, and successful academic work in majors that sparked passion and challenged their intellects. As seniors, they measured their academic success by those standards. As one student said, "What we learn in the classroom is just a fraction of what we are learning here." That fraction is not the only kind of college learning that matters to faculty, students or parents.

Second, national standardized tests would be too small a thimble even to capture students' academic learning. UW SOUL results show that college learning varies, depending on the academic discipline and, particularly, on students' majors.

It seems obvious that Astronomy majors leave college with different content knowledge than Electrical Engineering or International Studies majors. How-ever, skill-based learning, such as writing and critical thinking, also varies by discipline. An excellent Biology paper uses structure, methods, evidence and other practices quite different from those found in an excellent History paper. Furthermore, the thinking Art students use to critique their own and others' work differs from that Chemistry students use to analyze error in lab results, although both approaches might be labeled "critical thinking."

No single test can capture the practices, conventions and preferences of all disciplines equally and fairly. At best, such tests would prove useless to depart-ments seeking information to improve their curricula. At worst, they might end up reshaping curricula to serve the tests' own narrow ends.

Higher education needs meaningful assessment rooted in departmental learning goals and conducted by faculty who know the practices of their fields. We need assessment that recognizes the complexity, diversity and disciplinary nature of college learning-in other words, assessment that can offer a credible direction for improvement.