Home Articles Reader Opinion Editorial Book Reviews Discussion Writers Guide About TCRecord
transparent 13

But Can They Teach? In Praise of College Professors Who Publish

by Robert A. McCaughey - 1993

Suggests that highly scholarly active faculty exist in substantial numbers on select college campuses and that such faculty are more likely to be perceived as effective teachers than faculty who are not scholarly active. Presents some findings from an ongoing reseach project as an argument about faculty priorities. (Source: ERIC)

The current debate over faculty priorities—“teaching versus research” in its Manichaean form—is at a critical stage. Broad agreement now exists, as critics of research universities and the “research-university model” like William F. Bennett, Allan Bloom, Page Smith, Charles J. Sykes, Bruce Wilshire, and Lynne V. Cheney have been arguing for a decade, that the quality of undergraduate teaching on university campuses has suffered, if not, per Cheney, “as a by-product of an excessive emphasis on research,” then from protracted inattention.1 In calls to reform the way their own universities provide for undergraduate instruction, the outgoing presidents of Stanford and Columbia have taken such criticism to heart.2 Meanwhile, those defending the research university, notably Derek Bok and Henry Rosovsky of Harvard and Jaroslav Pelikan of Yale, have shown little inclination to insist, as Thorstein Veblen did so confidently in the early ascendance of the university, that teaching has importance “only in so far as it is incidental to an aggressive campaign of inquiry.“3

But if there is an emerging consensus that university faculty should attend more to teaching, especially undergraduate teaching, does it follow that the faculty of liberal arts colleges can—or should—attend less to scholarship, especially that intended for publication? Recent contributors to the faculty-priorities debate who have declared research and teaching to be “inescapably incompatible” logically argue that it does.4 Several critics of the research-university model, including recent National Endowment for the Humanities Chair Lynne V. Cheney, having earlier warned against the emergence of “Oberlin-led research colleges” and decried college faculty who distinguish between teaching and “my own work,” have implied as much.5 Although less vehemently, so have such warm advocates of liberal arts colleges as Ernest L. Boyer in regularly quoting those who regard the relationship of faculty scholarship and effective teaching as zero-sum.6

Heeding such advice would yield substantial economies for colleges. These would occur chiefly by reversing the recent trend toward lighter teaching programs and more generous faculty leave policies, moves explicitly intended to foster faculty scholarship.7 Similarly, funds now used to provide faculty with on-campus research facilities, electronic hook-ups to scholars elsewhere, and research-related travel could be put to more direct instructional purposes.

Noneconomic incentives also exist for colleges to downplay, if not actively discourage, the scholarly pursuits of their faculty. Elaborate procedures now in place on college campuses for externally assessing the scholarly credentials of faculty considered for tenure could be dismantled, allowing tenure to turn simply on a faculty member’s locally perceived teaching effectiveness and college service. Moreover, if, as Boyer and others suggest, college faculty much prefer teaching to pursuing research, getting out of the business of promoting faculty research and focusing single-mindedly on teaching would make colleges happier places.8

Finally, should colleges opt for a version of Nancy Reagan’s advice to kids tempted by drugs and “just say no” to faculty scholarship, public reaction is almost certain to be positive. Such has been the case since at least 1870 when Yale’s President Noah Porter launched the first preemptive attack on “professors . . . who have little or no active concern with the business of instruction or who come rarely in contact with the students.“9 The best press Robert Maynard Hutchins ever had was when he took on the research-oriented faculty of the University of Chicago on behalf of its “neglected” undergraduates.10 Thus, a single-minded commitment to teaching by their faculty, in sharp contrast with the widely reported “flight from the classroom” among their university counterparts, could only win colleges favor among government officials, editorial boards, and parents already openly contemptuous of “absentee faculty.“11 Page Smith, after all, meant it as a compliment when he recently contrasted universities, where “routine and pedestrian research” reigns, with “the hundreds upon hundreds of small, obscure colleges, denominational and non-denominational, whose faculty teach devotedly and whose students learn happily and well.“12

Such well-intentioned advice notwithstanding, I am persuaded that the institutional ends of select colleges that actively support the scholarly lives of their faculty will be well served by their continuing to do so. I am further persuaded that among select colleges where such support has been less active and more equivocal, increased support for faculty scholarship would advance their institutional ends.

The institutional-ends test used here is a rigorous one. It requires more than that faculty scholarship, as recent research concedes, “does not significantly detract from performance as a teacher”;13 it must be positively linked to effective teaching. Moreover, the linkage must not only be the institutional level—colleges with a quota of Morris Zapps offset by an equal number of Mr. and Ms. Chip’s—but at the individual level. This requires identifying specific faculty in appreciable numbers at select colleges who are both active scholars and effective teachers. Only then can it be argued that “scholar-teacher” is not an academic oxymoron trotted out by commencement orators, but, at least at the select colleges examined here, a reality that is alive, well, and worthy of institutional support.


The case for scholar-teachers on select college campuses logically begins with establishing the prior existence there of ”scholars.” This by no means goes without saying. Two important studies of the American professoriate during the past two decades, each based on extensive national faculty surveys, have stressed in their findings how few college professors ever publish something of scholarly consequence—indeed, publish anything at all. It was Everett C. Ladd, Jr., who first reported in 1979 that “most faculty don’t like research and don’t do it very well, if at all.” But it has been Ernest L. Boyer’s widely quoted findings published in l987 that 67 percent of all liberal arts college faculty he surveyed reported that they had never published a book, 38 percent had never published in a professional journal, and 49 percent were not currently engaged in scholarly research that will lead to publication that statistically renders college campuses scholarly wastelands.14

Other studies inspired by and drawing on Boyer’s work cheerfully concede, if not applaud, the absence of publishable scholarship produced on college campuses. Eugene Rice’s work-in-progress “The New American Scholar,” in its call for broadening the definition of scholarship beyond that operative at research universities—“the scholarship of discovery,” published in specialized journals—to include virtually all forms of intellectual activity, is a case in point.15 Kenneth P. Ruscio’s work, wherein he both attributes to and urges on college faculties “a distinctive scholarship” that is more interdisciplinary and less technical than that favored by their university-based disciplinary colleagues, is another.16

The significant differences between the findings of Boyer and Ladd and those reported here turn in part on differences in the composition and relative sizes of the institutional samples, as well as the resultant overall size of the sample college faculty populations. Their-sampling procedures yielded much larger numbers of faculty overall, but far fewer liberal arts college faculty.17 It is still the case, however, that the twenty-four colleges from which this sample draws its faculty are not representative of all 700-odd liberal arts colleges in the United States, or even of the 142 “Liberal Arts I” colleges as classified by the Carnegie institutional classification scheme. They represent instead a reasonably inclusive sample of those liberal arts colleges most likely to have within their faculties substantial numbers of women and men actively engaged in the scholarship of their respective disciplines.

Other sampling considerations included a desire for some measure of geographical dispersion, variety in enrollment characteristics, and a substantial representation of women among the sampled faculty (see Table 1). Colleges included also had to agree to participate in a study of the scholarly lives of college faculty. To be sure, considerable overlap exists between these “select” colleges and various listings of the “best” or “top” colleges, the best known being that published by U.S. News and World Report. There are important differences as well, the result of the use of three criteria here that play no part in the process U.S. News and World Report uses in coming up with its “America’s 25 Best Colleges”:18


“Select” colleges recruit faculty overwhelmingly from nationally ranked Ph.D. programs;19

their faculty are regularly among the winners of national faculty fellowship competitions;20

they rank among the leading AR-granting institutions in the proportion of their graduates going on to earn Ph.D.‘s.21

As yet untested hunches about the structural ways science faculty at select liberal arts colleges differ from both their science colleagues at research universities and their nonscience faculty colleagues led to their exclusion here.22 Those included are faculty in the humanities and social sciences as represented by the full-time membership of ten departments (see Table 2).

These institutional and disciplinary parameters yield a sample of 1,647 full-time faculty teaching at select colleges in-academic year 1991-1992. Readily observable sample characteristics include the following:

Almost a third (31 percent) are women.

One in five went to college at one of the sample colleges.

More than a third received their Ph.D.‘s from one of three universities—Harvard, Yale, Columbia.


Their average age (in 1992) was 44.

They have been teaching for an average of ten years.

Most (65 percent) have spent their teaching careers where they now teach.

Nearly two-thirds (63 percent) are tenured.

Identifying demographic and educational characteristics of the faculty sample is relatively straightforward, particularly with help from the colleges included. Identifying who among them are scholars—and who are not—is trickier. The procedure used took three steps, each labeled by the methodological question it is intended to answer.

Question 1: Who is a scholar?

Answer: An academic who engages in scholarly activity. “Scholarly activity” is intellectual activity, here narrowly defined as that which gets published primarily by university presses or in specialized journals for which it will have been previously reviewed and is likely to be read and later cited by disciplinary colleagues. “Disciplinary colleagues” are other scholars engaged in related scholarly activity, usually at institutions.

Question 2: How is scholarly activity measured?

Answer: By counting published instances of it. Whereas most earlier studies of faculty scholarly activity have depended on self-reporting, which requires faculty cooperation and must allow for exaggeration, reliance here is on an externally generated, discipline-specific composite of publication and citation counts. Publications include books (as measured by the number of individual listings in the electronic version of the OCLC “World Catalogue”) and articles (as measured by the number of articles listed in a given discipline’s most comprehensive bibliographical source). Citation counts were conducted using the Social Science Citation Index and the Arts and Humanities Citations Index, both published and updated regularly by the Institute for Scientific Information.23

Question 3: Just how scholarly active is “scholarly active”?

Answer: There are two different ways of measuring this. The first is by comparing the scholarly activity of an individual faculty member against the average scholarly activity of faculty in that discipline. Accordingly, if one is more scholarly active than is the average among one’s disciplinary colleagues, as indicated by one’s raw score of books, articles, and citations, one is scholarly active. When a faculty member’s discipline-specific raw score of scholarly activity is then standardized into what statisticians call a z-score (raw scholarly score—disciplinary mean scholarly score/the standard deviation of the disciplinary distribution of raw scores), comparisons across disciplines become possible. Thus, for example, a scholarly z-score with a positive sign (e.g., +.500) indicates scholarly activity above the average for the faculty member’s discipline; a negative z-score, below the average. The magnitude of the z-score locates a-faculty member’s scholarly activity in terms of standard deviations above or below the disciplinary mean. When aggregated by department, individual scholarly z-scores permit comparisons across disciplines within a given faculty or with counterpart departments elsewhere; institutional aggregations allow comparisons among sample colleges.

However useful for such purposes, scholarly z-scores only permit comparing one college faculty member—or one set of faculty members—with another. Such comparisons do not, for example, tell us how a scholarly active historian at a select college compares with another historian at a research university. For comparisons of this second sort a control group of university-based faculty is used. It consists of all 745 full-time faculty teaching in the ten sample disciplines at Columbia, Princeton, and Yale universities in academic year 1991-1992 (see Table 3).

Publication and citation counts done earlier for 1,647 college faculty were repeated for these 745 university faculty. Individual scholarly scores were then generated, which, when aggregated, permitted the establishment of median and mean “university levels” of scholarly activity for all ten disciplines. These levels became the second means by which the scholarly activity of select college faculty is measured. Thus, a college historian with fewer indicators of scholarly activity than obtained for most university-based historians (i.e., below the median number) is deemed “less active”; another, whose scholarly activity places him or her above the median of university historians, to be “as active”; and a third, whose scholarly activity exceeds the mean of all his or her university colleagues, “more active.” Here, as well, aggregation permits intra- and interinstitutional comparisons.



Both the analysis permitted when raw scholarly scores are standardized into z-scores and that when raw scores are compared with university-based colleagues point to the same conclusion: Scholarly activity among the faculty of these select colleges is substantially greater than earlier studies have indicated for liberal arts colleges generally. To be sure, such scholarly activity is unevenly distributed within and among faculties. In addition, the proportions of college faculty who perform at levels of scholarly activity comparable to their university-based disciplinary colleagues vary substantially from discipline to discipline and from college to college (see Table 4).

For all their differences, all twenty-four colleges employ some faculty whose scholarly activity places them well within the range characteristic of university faculties. Overall, the ratio of such “university-level” faculty to all faculty is about one in four. At the half-dozen colleges where such faculty is most concentrated, the ratio exceeds one in three. But even at the half-dozen with the fewest, where the ratio is more on the order of one in eight, the existence of scholars within their faculties may be, for present purposes, safely assumed.


If identifying scholars among college faculties where their very existence has been questioned presents one kind of methodological challenge, identifying “teachers” among faculties wholly engaged in teaching presents quite another. Moreover, if the above attempt to measure scholarly activity drew criticism for its inherent reductiveness, confusion of quantity and quality, anal-retentive manifestations, whatever, the following attempt to quantify effective teaching is sure to be viewed as blasphemy.24

Chief among those resistant to the notion-that teaching can be meaningly evaluated are the faculty themselves. Senior faculty particularly resist having their teaching evaluated, which some consider an infringement of academic freedom. Yet the institutionally perceived need to make such evaluations, and to do so in more systematic ways than faculty have traditionally depended on to evaluate the teaching of their junior colleagues, has prompted several select colleges to implement course-evaluation programs that produce data used in evaluating the overall teaching effectiveness of their faculty. Several others have such programs under active consideration. Others now systematically poll graduating seniors about the strengths and weaknesses of their various instructors, while still others utilize reports from advisers, departmentally produced submissions to tenure and promotion committees, and even student-administered questionnaires to assess teaching effectiveness.

What both faculty skepticism about evaluating teaching generally and the absence of uniformity in evaluative arrangements now in place point up is the particularistic nature of teaching and of its evaluation. It is not because select colleges or their faculties are indifferent to teaching that they have not produced authoritative standards for judging teaching; it is because they recognize that all teaching is local.

The means devised here to identify faculty by their teaching effectiveness accords with this reality. It consisted of request to the chief academic officers of the select colleges, usually the dean of the faculty or the provost, to rate a specified sampling of the college’s faculty on a relative scale of 5 (“one of the college’s most effective teachers”) to 1 (“not regarded as among the college’s more effective teachers"). In doing so, they were urged to conform to the law of averages, so that the number of faculty given the highest ratings was matched by an equal number of faculty assigned low ratings. The faculty for whom ratings were requested were drawn primarily from the ranks of the college’s most and least scholarly active members, although they were not identified as such. In addition, they had to have been teaching at their present colleges for at least five years.

To date, the chief academic officers of seventeen select colleges have provided teaching ratings on 512 of their faculty. In most of the statistical manipulations that follow, those with ratings of 3 (“one of our typically effective teachers”) have been dropped to allow dichotomization of the population into those counted among their college’s “most effective teachers” (5s and 4s) or among their “less effective teachers” (1s and 2s). Of the 345 select college faculty so rated, 240 are in the first category—hereinafter “Good Teachers”—and 105 in the second—for our dichotomous purposes, “Not So Good.”

Deans reported a variety of means used to arrive at the assigned ratings. Some drew directly on collegewide teaching-evaluation systems for ratings that could be aligned with the 5 to 1 (“yeah” to “nay”) scale of the study. Others depended on both quantitative and qualitative information available from the college’s faculty review and compensation committees. Several deans checked their ratings with previous deans. Although anecdote and rumor undoubtedly played some part in arriving at these ratings, deans primarily dependent on little else did not provide teacher ratings.

Several findings that merit comment emerge from these ratings. Some support widely held assumptions about teaching effectiveness, others do not. It does appear, for example, that some disciplines (e.g., Classics and History) are more generally credited with effective teaching than are others (e.g., Philosophy and Economics), independent of institution (see Table 5).25

Similarly, as other studies have suggested, teaching effectiveness and age among select college faculty are inversely correlated, independent of discipline and institution (see Table 6).26 Contrary to some recent studies, however, women faculty is here perceived to be marginally better teachers than their male counterparts (see Table 7).27




What is central here is that by establishing a defensible means by which select college faculty can be classified into “Good Teachers” and “Not So Good,” it becomes possible to assess the correlative relationship between “Teaching Effectiveness” and “Scholarly Activity,” with its heretofore dichotomously arrayed “Scholars” and “Abstainers." Let us proceed directly to do so.


While on college campuses the known variety of academic types approaches infinity, the permutations here, limiting ourselves to two variables, each of which is dichotomous, are only four:

1. “Scholars”—those faculty whose scholarly activity exceeds that of their disciplinary colleagues, but whose perceived teaching effectiveness falls short of institutional colleagues.

2. “Teachers”—those faculty whose perceived teaching effectiveness exceeds that of their institutional colleagues, but whose scholarly activity falls short of that of disciplinary colleagues.


3. “Neither’‘—those faculty whose teaching effectiveness is less than that of their institutional colleagues and whose scholarly activity is less than that of most of their disciplinary colleagues.

4. “Scholar-Teachers”—those faculty whose scholarly activity exceeds that of their disciplinary colleagues and whose perceived teaching effectiveness exceeds that of their institutional colleagues.

These possible permutations can be visualized through the use of a 2 x 2, or four-celled cross-tabulation, wherein the two categories for “Teaching Effectiveness” are arrayed in rows, the two for “Scholarly Activity” as columns (see Table 8).

Faculty with paired characteristics that place them in the first and fourth cells support the hypothesis that scholarship and teaching are negatively correlated, incompatible, pursued at cross purposes; those paired so as to be placed in the second and third cells support the hypothesis that scholarship and teaching are positively correlated, compatible, interdependent. A third hypothesis, the null hypothesis, that there is no relationship between scholarship and teaching, would be supported if faculty distributed themselves among all four cells in proportion to the respective proportions within two variables (i.e., the “expected value”).

As shown in Table 8, faculty do not distribute themselves randomly, so the null hypothesis that no relationship exists between scholarship and teaching may be set aside. Furthermore, insofar as both the first cell (“Teacher”) and the fourth (“Scholar”) have lower counts than would be expected of random pairings, it can be concluded that such “one-but-not the-other” pairings are not the norm, which in turn allows setting aside the hypothesis that scholarship and teaching are inversely related.

That leaves the second cell (“Scholar-Teacher”) and the third (“Neither”), both of which have higher counts than predicted by the null hypothesis. Thus, both support the conclusion that among the sampled select college faculty there is an overall correlation between scholarship and teaching, that it is positive, and that it is statistically significant (rø= +.1965, +.1965, p<.01).

Although the numbers of faculty involved are often too small to assure statistical significance, tests for correlation between scholarship and teaching at the institutional and disciplinary levels produce consistently positive results. Similarly, the correlation is positive-for both women faculty (rø= +.1988, p<.08) and men (rø=+.1951, p<.01) (see Tables 9 and 10).

But perhaps the most tantalizing finding relates to the strength of the scholarship-teaching relationship among the various age cohorts of the faculty sample. Far and away the most positive correlation occurs among the eighty-one teaching-rated faculty age sixty and older (rø= +.35, p<.01). When combined with the earlier finding that teaching effectiveness is inversely correlated with age, this provides substantial support for a view espoused in several interviews of scholarly active senior faculty: that their effectiveness in the classroom had been sustained—or at least prolonged—by their ongoing scholarly activity. To the extent that this is so, institutional support of faculty scholarship should be viewed as an increasingly important investment in sustaining teaching effectiveness as colleges enter the era of no mandatory retirement.




This article makes two main points: that highly scholarly active faculty exist in substantial numbers on select college campuses, and that such faculty are more likely to be perceived as effective teachers than faculty who are not scholarly active. A third logically follows: that scholarly abstinence among faculty of select colleges does not promote good teaching. From these flow a policy-related inference: that select colleges assure teaching excellence, their traditional raison d'être and increasingly recognized niche in the highly differentiated American higher education system, by actively supporting the scholarly lives of their faculty.

Even if one accepts the argument that select colleges should support scholarly activity in the interests of good teaching, the causal question remains: Why is it that scholarly active college faculty make good teachers? That is another article. I would suggest, however, that an essential source of insight into the causal relationship between scholarship and teaching is the scholar-teachers this paper both identifies and celebrates.

This paper presents a preview of an ongoing research project as an argument about faculty priorities. Support for the “Liberal Arts Colleges and the Higher Learning” project has been provided by the Spencer and Andrew W. Mellon Foundations. Proceedings of the Conference of Select Colleges held at Barnard College, January 23-24 1992, at which several themes of the project were discussed by representatives of twenty-one colleges, scholars, foundation officials and journalists, are available from the author.

Cite This Article as: Teachers College Record Volume 95 Number 2, 1993, p. 242-257
https://www.tcrecord.org ID Number: 79, Date Accessed: 5/20/2022 10:57:41 PM

Purchase Reprint Rights for this article or review
Member Center
In Print
This Month's Issue