Grading the Effects of a Liberal Arts Education
David G. Winter et, al
For more than 2.000 years, a liberal education has been the ideal of the West—for the brightest, if not for all, students. The tradition goes back to Plato, who argued in The Republic that “leadership should be entrusted to the philosopher—“a lover not of a part of wisdom only, but of the whole… able to distinguish the idea from the objects which participate in the idea.” More recently, in a World War II-era treatise, a Harvard University committee concluded that a liberal education best prepared an individual to become “an expert in the general art of the free man and the citizen.” The report, which led to the introduction of Harvard’s general education curriculum, concluded, “The fruit of education is intelligence in action. The aim is mastery of life.”
In recent years, the fruit has spoiled and such high-sounding rhetoric has been increasingly challenged. Critics have charged that liberal arts education is elitist education, based on undefined and empty shibboleths. Caroline Bird, social critic and author, argues in The Case Against College that the liberal arts are a religion, “the established religion of the ruling class.” Bird writes, “The exalted language, the universalistic setting, the ultimate value, the inability to define, the appeal to personal witness…these are all the familiar modes of religious discourse.”
Students in the 1960s charged that such traditional liberal arts courses as “Western Thought and Institutions” and “Contemporary Civilization” were ethnocentric and imperialistic. Other students found little stimulation in a curriculum that emphasized learning to both formulate ideas and engage in rational discourse. They preferred, instead, to express themselves in experience and action; they favored feeling over thought, the nonverbal over the verbal, the concrete over the abstract. In the inflationary, job-scarce economy of the 1970s, many students argue that the liberal arts curriculum is “irrelevant” because it neither prepares them for careers nor teachers them marketable skills. In its present form, moreover, liberal arts education is expensive education.
Partly in response to these charge and, more immediately to faculty discontent, Harvard recently approved a redesigning of the liberal arts program. Faculty had complained that the growing numbers and varieties of courses had “eroded the purpose of the existing general education program.” Students, they felt, could use any number of courses to satisfy the university’s minimal requirements, making those requirements meaningless. The new core curriculum will require students to take eight courses carefully distributed among five basic areas of knowledge. The Harvard plan proposed to give students “a critical appreciation of the ways in which we gain knowledge and understanding of ourselves.” Plausible as this credo may be, it rests on rhetoric and not solid research evidence—like curriculum innovations of the 1960s.
In an era of educational accounting and educational accountability, it would be helpful to have a way of determining what the essential and most valuable “core” of a university education is and what is peripheral and mere tradition. What are the actual effects of a liberal education, this most persistent of Western ideals? It is sobering to realize that we have little firm evidence.
Against this background, we recently designed and carried out a new study to get some of the evidence. Our findings suggest that liberal arts education does, in fact, change students more or less as Plato envisioned, so that the durability of this educational ideal in Western civilization may not be undeserved. In our research, liberal education appears to promote increases in conceptual and social-emotional sophistication. Thus, according to a number of new tests we developed, students trained in the liberal arts are better able to formulate valid concepts, analyze arguments, define themselves, and orient themselves maturely to their world. The liberal arts education in at least one college also seems to increase the leadership motivation pattern—a desire for power, tempered by self-control.
The precise content of a liberal education remains unclear. Is it the study of certain “core” disciplines or bodies of knowledge—courses in Western civilization or modern literature, or a particular set of “Great Books”? Does it require a multidisciplinary approach, as, for example, in courses entitled “Science and Responsibility” or “Freedom and Authority in the Modern Novel”? Many professors argue that the essence is not what is learned, but how it is taught—with an emphasis on concepts rather than facts, on independent inquiry rather than learning by rote. Some educators, perhaps half-facetiously, contend that liberal arts include everything that is not of obvious practical or vocational use!
The Harvard committee during World War II theorized that general education fostered four traits of mind: thinking effectively, communicating thought, making relevant judgments, and discriminating among values. Some 33 years later, the committee headed by present dean Henry Rosovsky characterized the goals of the liberally educated person in similarly luxuriant language: to “think and write clearly and effectively”; to have “some understanding of, and experience in thinking about, moral and ethical problems”, and to use experiences in the context of “other cultures and other times.”
Still, these traits and skills remain largely unmeasured and ignored by psychologists who, even when they study “thinking,” focus on much more elemental and simple processes. Most of the abundant research on the effects of higher education have focused on changes in personality, values, and beliefs. Even here, the conclusions are largely equivocal: many college “effects” are to the process by which the students were chosen in the first place and not to the changes that occur during college. Studies have shown, for instance, that attitudes are stabilized as much as they are altered during college.
We started our study from two fundamental premises: first, that the evidence to date was probably more a reflection of the testing procedures used than of the efficacy of higher education; and, second, that new tests should be modeled on what university students actually do rather than on what researchers can easily score. If liberal education teaches articulate formation of complex concepts, then student research subjects should be asked to form concepts form complex material and then scored on how well they articulate them, rather than being asked to choose the “best” of five concepts by putting a check mark in one of the boxes. In more formal terms, tests of the effects of education should be operant tests that require operating on material and making up answers, rather than respondent tests that merely ask for choices from among preceded alternatives.
Any study of the effects of higher education has the difficult task of distinguishing educational effects from simply maturational effects. In order to have some control over the effects of maturation, therefore, we tested students who were receiving three different kinds of higher education:
1. A traditional four-year liberal arts education at a prestigious Eastern U. S. institution. By any definition, students attending this school enjoy a curriculum that is considered liberal arts. It is a well-endowed, private college with a tradition of scholarly excellence, an eminent faculty, and great prestige. Its students, drawn from this country and abroad, must satisfy very competitive admissions standards. The college accepts 20 percent of all applicants. Approximately two-thirds of its students are men and one-third women. The curriculum emphasizes broad, interdisciplinary survey courses in the sciences, humanities, and social sciences, and individualized scholarship at all stages of the college career.
2. A four-year undergraduate program for training teachers and other professionals. The offerings at this state-controlled institution have been expanded in recent years to include such general and career programs as law enforcement and health education. The college’s students, drawn from a large metropolitan area, must pass moderately competitive admissions standards; about one-half of those who apply are accepted. The students body is about evenly divided between men and women.
3. A two-year community college that offers career programs in data processing, electronics, nursing secretarial skills, and business administration. A publicly controlled institution, our community college is situated in a city and draws most of its students from nearby suburbs. It has a relatively nonselective admissions policy, accepting about 70 percent of those who apply. The student body is 60 percent male.
We administered three kinds of tests to a total of 414 students, half men and half women, drawn from the first-year and last-year classes of the three colleges. We controlled statistically for intelligence and social class, to eliminate differences in performance based on these two characteristics. By comparing the test scores of first-and last-year students at each school, we hoped to determine the degree and nature of any changes brought about by the educational programs. By evaluating all three schools together, we hoped to find out whether the liberal arts school has unique impact on its students.
With our new Test of Thematic Analysis, we examined the students’ abilities to create and express sophisticated concepts. We asked them to read two groups of brief, imaginative stories and then to describe the differences between the two in any way they liked. We awarded positive values to their work when they perceived characteristics of both story groups that could sensibly be compared and contrasted, used examples and qualifications to strengthen their arguments, legitimately redefined aspects of stories to support their theses, and found general categories to group apparently unrelated elements. When they compared unlike things or used affective and subjective phrasing such as “It makes the reader nervous” or “It left me satisfied,” we awarded negative values.
At all three institutions, last-year students scored higher than first-year students, but seniors at the liberal arts college far outdistanced their counterparts at the teachers’ and community colleges.
Thus, a typical freshman at any of the schools might describe the differences between the two groups of stories in rather wandering terms: “Group B stories are more exciting than Group A stories. They were about nasty leaders and I don't like that. Group B stories show people as not trusting each other.” A typical final-year student at our liberal arts college might put what is essentially the same contrast in these terms: “Both groups of stories involve relations to authority. In Group A, authority is either accepted or actively rejected; while Group B stories involve moderate suspicion of authority. While story A-4, an animal fable, might seem an exception to the rule, it does, in fact, fit if one considers the phrase ‘king of the beasts’ as representing symbolic authority.”
Liberal education, then, seems to affect the way in which people marshal, organize, and “operate” on facts. These processes are spontaneous, self-initiated, and active, and are the same ones called for by an essay assignment to “compare and contrast the Renaissance and the Reformation,” or an examination question asking, “What are the essential differences between normal and malignant cells?” We believe these processes are more central to a liberal education than learning simple concepts and memorizing detailed facts. Indeed, we gave the students an adaptation of a standard reading-comprehension test and found that none of the three schools significantly affected the ability to learn and remember isolated facts.
As another means of probing the conceptual processes and reasoning abilities of the three groups of students, Abigail Stewart devised an Analysis of Argument test. The test first quotes an extreme, unpopular, and rather badly argued position on a controversial issue. Students must attack this position and support their own stance with reasoned argument. Again, the quality of the attacks improved from first to final year at all three schools, but more so at our liberal improved from first to final year at all three schools, but more so at our liberal arts institution. Thus, a typical first-year student would dispute a series of facts: “X is wrong when he says that…” A final-year attack focused on a more abstract, general principle, such as faulty logic: “X’s arguments all derive from a confusion of association with causation.”
In the second step of Stewart’s test, students switched sides and had to defend the position they had formerly attacked. Most floundered and simply substituted a blank endorsement for a blanket attack. Only our final-year liberal arts students were able to craft a limited, qualified endorsement of a position they had opposed. They could respond: “While there are flaws in X’s whole line of reasoning, it must be admitted that some of his particular claims and examples are true.” In other words, the liberally educated students were better able to argue both sides of a question, but with integrity and intelligence rather than by simply espousing the other point of view uncritically.
The other changes in student ability unique to, or more pronounced at, the liberal arts college involved measures in the Thematic Apperception Test (TAT), a “projective” test that clinicians have used for over 40 years to assess personality. Subjects see a series of vague and ambiguous pictures—a man wearing the uniform of a ship’s caption talking with a man in a business suit, for example, or two women in lab coats using equipment such as a test tube—and must tell or write stories about the pictures. Researchers may use any number of scoring systems to analyze the results, depending on the personality characteristics that interest them. We were looking for three elements in the responses; self-definition, maturity of adaptation to the environment, and the leadership motivation pattern.
A story that scores high in self-definition uses causal words such as “because” and “in order to,” and portrays characters who take actions for reasons, for example: “After being miserable for a while, the woman in the picture will realize that her love affair won't and will leave.” Low-scoring stories portray ineffective actions, events with no apparent causes, and characters who experience intense feelings in response to others’ actions, but who are unable to act themselves: “The man and woman pictured will try desperately to establish a love relationship, but will end up feeling only more alone.” In a number of studies, people who score high in self-definition act instrumentally (that is, effectively and constructively), often in ways that go beyond ascribed roles. Self-defining women, for example, tend to seek careers as well as marriage, and in many different ways are not limited by traditional sex roles. Thus, self-definition is associated with an instrumental, effective style of translating though into action. When compared with the teachers’ and community colleges, that liberal arts colleges produced unique and significant gains in student-self-definition.
Maturity of adaptation to the environment refers to success in developing characteristics that personality theorists have identified as representing the highest levels of personality growth or maturity. Drawing on the ideas of Freud and Erik Erickson, who described the “stages” of development, Steward recently worked out a TAT measure of this adaptation. As we expected, students at all three institutions showed higher stages of adaptation over time, but those at the liberal arts college showed larger, more significant gains. In terms of particular scoring categories from the Stewart measure, this means that students, our liberal arts students in particular: (1) see authority in complex, versus simplistic, pro and con terms; (2) view other people as differentiated beings in their own right, rather than as simple means of gratifying their (the students’) desires; (3) integrate both joy and sorrow into their moods; (4) are able to work without falling victim to passivity, self-doubt, or anxiety about failure.
Seniors at the teachers’ and community colleges scored higher than freshmen in both maturity of adaptation and self-definition, suggesting that almost any kind of higher education, or even just physical and social maturation, has some influence on these variables. But for all measures, the gains at the liberal arts college were significantly greater.
It appears that the liberal arts college also fosters a unique pattern of motivation in its students: strong concern for power and weak concern for affiliation, combined with high self-control or ability to inhibit activity. Thus, the final-year liberal arts students wrote more TAT stories with the following combination of characteristics: (1) one character has an impact (or tries to) on another; (2) activity is restrained or inhibited—as indicated by the use of such words as “not” or “cannot”; and (3) characters do not show concern with establishing and maintaining warm, friendly relations with others.
David McClelland, of Harvard, has called this set of characteristics the leadership or “imperial” motive pattern. In a series of experiments, McClelland has demonstrated that it is usually found n individuals who are considered effective leaders—managers who have a talent for creating in their subordinates such qualities as high morale, a sense of responsibility, organizational clarity, and “team spirit.” For, while the qualities of an imperial motive suggest that a person is not compassionate, they generally dictate that he will be fair, treating others in an impartial manner that subordinates seem to appreciate.
The preset study of only three colleges limits inference and further speculation. We must study other liberal arts schools to discover whether they have the same impact on the students as the liberal arts college discussed here. The issue will likely be complex. Indeed, data we recently collected from another college similar to the liberal arts school we examined suggest that liberal education there increases sell-definition, but decreases maturity and has little or no effect on the imperial pattern.
When we know more about what causes the kinds of changes in students detailed here, then our research can contribute to shaping educational policy. But who can say, from the evidence now at hand, that the effects of liberal education at our liberal education at our liberal arts college, or anywhere else, are caused by course requirements at all? It may be that the worth of an education at any school is determined more by faculty quality, library facilities, the size of the endowment, or even by the self-fulfilling anticipations and beliefs of faculty and students. We are currently seeking answers to these questions, taking our new test procedures to students at more than 15 different post-secondary-school institutions. During the next year or two, we hope to point to specific qualities of liberal arts colleges that leave their particular imprints on the students.
Still, the changes unique to, or enhanced by, attendance at our liberal arts college do establish at least a prima-facie case for education in the liberal arts. The pious goals and extravagant language of liberal arts educators must yet be analyzed, broken down into specific skills. With tests to measure student abilities in these skills, we can determine whether liberal arts education is doing what its proponents claim and how its performance can be improved.
