Association of American Colleges and Universities, January 2001

prepared by Ross Miller, AAC&U

Abstract

Plans for standardized testing of college students in public institutions appear to be on the rise. A proposal for testing students prior to graduation in the University of Texas System was made recently by one of the Texas regents. The idea met with faculty skepticism and pointed questions about the fairness or desirability of a standardized testing program for college-level outcomes. (See the October 6, 2000 Chronicle of Higher Education, pp A35, A38).

While most stakeholders in higher education are deeply interested in improving student learning outcomes, there is disagreement about whether or to what extent standardized testing programs will contribute to this goal. Specific teacher characteristics, selected pedagogies, graded homework, and tests that inform subsequent performance have all been linked to better student achievement. Standardized testing practice inherently limits the extent to which any test can inform student performance, given such factors as national scale, secure test contents, and non-local scoring.

Testing programs must surmount significant difficulties including deciding what to measure, allowing for differences in students and resources at each institution, creating and/or selecting tests, and motivating students to do their best work on the tests. Currently, the number of standardized tests available for assessing the general outcomes of higher education is very limited, with only four instruments used in most of the testing situations.

Brief information about several testing programs and the aims of testing follows (gathered from various sources). A glance at the knowledge and skills tested in state-wide assessment programs provides one more data point in investigating the common aims and purposes of college education.

This test consists of four subtests: essay, English language skills, reading, and mathematics. Skills were chosen by faculties of community colleges and state universities through the College-Level Academic Skills Project. Since 1984, students in public postsecondary institutions in Florida have been required to demonstrate achievement of the skills to be awarded an associate in arts degree and/or for admission to upper-division status in a state university in Florida. (more at www.firn.edu/doe/sas/clsthome.htm)

TASP must be taken by all students prior to beginning classes at any public community college, technical college, or university in Texas. Reading, writing, and math skills at the college entrance level are tested. Admission to a school cannot be denied on the basis of TASP scores, but students cannot take college credit courses until they have at least taken TASP. If their scores are low, remedial study is expected. TASP must be passed prior to earning a two-year degree or before enrollment in junior- and senior-level courses at a four-year school. It is generally recommended that students take this test during high school. High scores on several other tests (ACT, SAT, etc.) may be substituted for a passing TASP score. (more at www.thecb.state.tx.us/divisions/univ/tasp/hspage.htm)

Since 1972, students in the University of Georgia system have had to pass a test of reading and writing prior to graduation. The test must be taken, at the latest, upon completing 30 semester credit hours. If not passed by the time 45 hours have accumulated, students must then enroll in remedial courses until they have passed both parts of the test. (more at www2.gsu.edu/~wwwrtp/overview.htm)

Success in meeting state-defined goals brings financial rewards to state schools in Missouri and Tennessee. By documenting successful performance on selected student outcomes and other institutional measures, campuses earn additional funding. In Missouri, goals include increased minority graduation and better performance of all students in both general education and the majors. Locally-controlled programs are funded to support innovations in teaching and learning. In 1997, performance funding represented about three percent of the total operating budget for Missouri schools. In Tennessee, campuses also may develop their own projects including such things as strategic planning and improvements related to accreditation. With a focus on improving the quality of instruction, funds may be used to develop new programs and purchase instructional equipment. In both states the choice of a test is left up to each campus, but a nationally-normed test must be used if available. Scores are not a barrier to graduation for individual students, rather, institutional scores are reported along with other data for performance funding. In Tennessee, a school may qualify for an additional 5.45 percent of their operating budget.

College BASE—Tennessee

College BASE is criterion-referenced. It is designed to measure three cognitive processes (interpretive, strategic, and adaptive reasoning) within four content areas (English, mathematics, science, and social studies). The long form consists of 180 multiple-choice questions. A short form is also available covering only two of the four subject areas (English and mathematics). There is an optional forty-minute writing exercise.

The purpose of this test is not to report student ranks, but to diagnose the strengths and weaknesses of students and curricula. Two forms of the test are available so that pre- and post-testing may be administered for value-added assessment. (This description of College BASE adapted from www.santacruz.edu.)

ACT-CAAP—South Dakota, Arkansas, Wisconsin, Missouri, and Utah

Developed in 1988 by ACT, the CAAP includes six tests in the areas of mathematics, writing, reading, critical thinking, and science reasoning. There are two writing tests to choose from; one is in a multiple-choice format, while the other requires the student to write two twenty-minute essays. The other four area tests are forty-minute multiple-choice tests. The CAAP is norm-referenced and provides an institutional summary along with individual student reports. CAAP scores can be compared to scores predicted by ACT admission scores to estimate "value added" between admission and the time of CAAP testing. (ACT-CAAP description from http://planning.ucsc.edu/irps/stratpln/ge_assmt.htm.)

ACT-COMP—Tennessee

Designed as a measure of skills and knowledge gained through "general education," this test will soon be discontinued. COMP utilizes a unique format of two correct answers for each multiple-choice question. Scores are reported for 1) functioning within social institutions, 2) using science, 3) using the arts, 4) communicating, 5) solving problems, and 6) clarifying values. Establishing a link between the six scores and specific parts of the college curriculum, proved impossible in this writer's experience administering COMP at a small college.

Commentary

In a draft paper on statewide testing in higher education, Peter Ewell expresses a belief that new college-level tests may soon be developed as a spin-off of assessment trends in K-12 education. Demands for accountability and high standards in the public schools have led to new ideas in testing and assessment that have yet to be used widely in higher education.

At the same time that campuses are promoting innovative programs to advance sophisticated learning in general education and the majors, state legislatures may be moving in the direction of requiring assessment of much less sophisticated outcomes. Higher education could soon be subjected to the same "teach to the test" syndrome that public school teachers in high-stakes testing situations now experience. If a widely shared understanding of the aims and purposes of undergraduate education can be created, perhaps higher education can avoid this dilemma.