A significant change in any component of a stable business model is likely to have significant impact on all other components of the model. In an earlier post, I discussed the business model for higher education (How can we think about the new wave of innovations in higher education?), and pointed out that the profusion of off- the-shelf new college courses (NCLCs) such as MOOCs and open courseware provide an important new resource in higher education. It is not surprising that this new resource has the potential to shake up the traditional business model significantly, and in a series of three posts (How a course-rich world might impact higher education: I, II, III) I considered some of the changes that this resource might produce. In this post, l continue that discussion by considering how this new resource of NCLCs raises some important issues related to one of the core principles of the process component of the higher education business model: Academic Freedom.
The most commonly quoted rational for academic freedom comes from the AAUP 1940 Statement of Principles of Academic Freedom and Tenure. That statement begins:
The purpose of this statement is to promote public understanding and support of academic freedom and tenure and agreement upon procedures to ensure them in colleges and universities.Institutions of higher education are conducted for the common good and not to further the interest of either the individual teacher or the institution as a whole. The common good depends upon the free search for truth and its free exposition.
Academic freedom is essential to these purposes and applies to both teaching and research. Freedom in research is fundamental to the advancement of truth. Academic freedom in its teaching aspect is fundamental for the protection of the rights of the teacher in teaching and of the student to freedom in learning.It carries with it duties correlative with rights.
For understandable reasons, this concept of academic freedom has lead to a very general acceptance of the idea that the curriculum of an institution "belongs" to the faculty - they use their professional expertise to define the curriculum broadly, and to maintain overall excellence in implementation.
However, the next-to-last sentence in the AAUP description above has a somewhat hidden asymmetry that is put into greater evidence by the arrival of the NCLCs. Because the faculty control the curriculum, the student's freedom is the freedom to learn what the teacher who has freedom to teach is teaching. That obviously made sense when the only courses available to the student were those being presented by their faculty. But now, with the arrival of the MOOCs, open courseware, courses produced by such companies as Pearson, etc, there is an enormous range of courses easily accessed by any student. Is the asymmetry of learning and teaching contained in the AAUP guidelines still appropriate?
I discussed an example of the potential conflict that can arise from the addition of NCLCs to the resource mix in an earlier post, Why MOOCs threaten academic freedom -too much value for the student. That post was based on a very articulate and honest statement made by Chandrakant Panse, a professor of microbiology at MassBay (Community College) and president of the union chapter there. He believed that the MITx course that MassBay was experimenting with was of much greater value to the students in their job searches than was a similar MassBay course. Ultimately, he feared that this could lead to administrators requiring that he follow MITx courses because students will demand it. This would present a threat to his and other faculty's academic freedom - which is to teach what they think is appropriate. This is a realistic concern, but this application of academic freedom seems quite possibly to be in contradiction to the second sentence to the AAUP statement, which declares that academic freedom should be supportive of the "common good", and "not to further the interests of ... the individual teacher". If the students are better off taking the MITx course based on knowledge gained and on employer interest, it would seem that the common good would best be served by the course of action that Prof. Panse fears despite the changes it would bring about in faculty role.
It is clear that much of the generally negative faculty reaction to the MOOCs is based on extensions of reasoning such as that of Prof. Panse - the role of the faculty is diminished by the introduction of material from other sources that defines the direction of the course, thus decreasing the professors autonomy in violation of their academic freedom.
Another major component of the negative reaction to the MOOCs comes, however, from faculty who argue that the premise above is incorrect - students will not actually be well served if they are taught by the MOOCs, whatever the students themselves may think. According to this view, MOOC-taught students will learn less, or less thoroughly and broadly, and employers will ultimately look poorly on the courses. Consequently, in this view the common good will not be served by introduction of the MOOCs, so there is no conflict between the academic freedom of the faculty to reject the MOOCs and the service of the common good. The open letter from the San Jose State Philosophy Faculty provides a very articulate argument for this viewpoint.
So which of these viewpoints is correct? Are students having their educational opportunities inappropriately restricted by the academic freedom of the faculty, or is the academic freedom of the faculty enabling them to prevent bad educational choices being made by students and administrators?
The difficulty in resolving these two contradictory viewpoints is that arguments on both sides generally are being made based on personal conviction, not data. The critical question seems to me to be: are the educational outcomes achieved using MOOCs and other NCLCs as good as or better than would be obtained from the traditional classes that they replace? Very little data that is really pertinent to answering this question exists. Most often, arguments about this subject involve comparing what most faculty firmly believe will be the outcomes obtained with NCLCs to outcomes of the ideal residential collegiate educational experience; the actual reality of the student learning experience at the institution involved is almost never part of the conversation. In fact, many of the conversations about the status quo could lead one to believe that it is so wonderful that every undergrad in the country currently must have his or her own log, with a faculty reincarnation of Mark Hopkins at the other end.
So what is needed at this point are some real experiments that compare outcomes of classes taught with NCLCs to outcomes of classes taught in traditional ways. The best way to start is with institutions considering adopting the NCLCs in some way. There are many ways to design such experiments, of course, but the relatively recent meta analysis by the Department of Education of comparisons of effectiveness between online and traditional instruction used a minimally necessary set of criteria:
- students should be assigned at random to the two types of classes teaching similar material
- measure effects of only objective measures of learning (e.g. not perceptions).
This DOE analysis of online vs traditional classroom teaching reached many conclusions, including:
Students who took all or part of their class online performed better, on average, than those taking the same course through traditional face-to-face instruction.
The effectiveness of online learning approaches appears quite broad across different content and learner types
This DOE study, although helpful and suggestive, can't really answer the question posed above. That question asks not about the general issue of online vs in class instruction that was considered by this study (already a difficult question), but whether use of a MOOC or other NCLC can lead to better learning outcomes in specific classes than similar classes where course material is defined and taught in a traditional manner by traditional faculty.
The answer to this likely will depend in a broad way on two sets of criteria. The first relates to the existing traditional class, and the actual current effectiveness of learning in that class. Data in general are not too encouraging on this front (e.g. Bok, Our Underachieving Colleges), but better and more complete learning measures are needed and they need to be more broadly used in order to make the needed comparisons. One can expect that some types of institutions and some individual faculty will show greater effectiveness in the traditional classroom than others, so the the standards to be used for comparison will likely vary by type of institution and even by faculty. The second relates to the effectiveness of the specific NCLC being used. As discussed in the three part series referenced above, the pedagogical quality of the NCLCs varies from extremely poor to very sophisticated, with some also incorporating such personalizing elements as adaptive learning. There are also many ways in which the NCLCs can be used. And, of course, the effectiveness of many of the NCLCs is increasing rapidly with time.
By the minimalist experimental criteria above, the recent the Udacity/San Jose State partnership does not qualify as a useful experiment in the use of NCLCs, and the visible failure of this partnership has simply increased the volume and strength of the belief-based convictions being expressed by both sides of the discussion. As such, it should provide a useful warning of the dangers of poorly designed and executed experiments in areas with strong competing philosophical positions.
Will these experiments get done? They are badly needed to help form effective educational strategies and policies in these times of restricted resources, with the forces of globalization increasingly demanding that we improve the educational outputs of our system. But the track record for doing the experiments thus far is not encouraging. Faculty have generally used their academic-freedom-based control of the curriculum to derail experiments on the needed scale. Their argument is most often that use of the NCLCs will be a disservice to the students because it will lead to an inferior education - essentially arguing that there is no reason to do the experiment because they know the answer already without supporting data. This purely faith-based argument has thus far been quite effective.
But some of the arguments made to go beyond the simple statement that the NCLCs will lead to an inferior education, and move into territory that leads us back into to the question of whether academic freedom is being used to advance the common good in this arena. For example, the San Jose State letter that argues against doing an experiment to look at the effectiveness of a MOOC goes beyond the argument about effectiveness to state:
We believe the purchasing of online and blended courses is not driven by concerns about pedagogy, but by an effort to restructure the U.S. university system in general......
....With prepackaged MOOCs and blended classes, faculty are ultimately not needed.
However understandable these concerns may be, this argument would seem to conflict directly with the AAUP "common good" statement:
Institutions of higher education are conducted for the common good and not to further the interest of either the individual teacher or the institution as a whole
since the argument seems to be that the new developments should not disturb the current structure and operations of universities or threaten the interests of faculty by putting their jobs at risk.
Surely, the best interests of the students should be a key (if not the key) component of the common good - providing the best possible education to students is what higher education is (or should be) all about. We need to see some of these experiments move forward so that we can all get a clearer picture of how the common good can best be served.