Research and Teaching: Conditions for a positive link [1]
Teaching in Higher Education, 6 (2001), pp 43 – 56.
Lewis Elton
University College London
Abstract
It has become increasingly clear over the past decade that the question of a positive link between research and teaching has no simple or general answer. At the same time, there may well be a positive link under particular conditions. This paper argues that a positive link can be due primarily to the processes, rather than the outcomes, inherent in research and teaching and that, in particular, student centred teaching and learning processes are intrinsically favourable towards a positive link, while more traditional teaching methods may at best lead to a positive link for the most able students, who in the perception of traditional academics are of course the future university teachers. This finding in turn leads to a rational explanation of the persistent myth of a general positive link. Finally, it is argued that pedagogic research and its outcomes could play an important role in strengthening the link.
Introduction
In the past decade, discussion of and research into the question of a link between research and teaching has proliferated. Although in the process much of significance has been learned, general insights have only rarely emerged. The purpose of this paper is not to add to the primary research in the area, nor to analyse earlier research, but to obtain a better understanding through reflection on what has emerged from other studies, through a combination of conceptual clarification and an argument for a specific position (see 1. below), with the conceptual clarification as a continuous thread in the paper. No attempt will be made to comprehensively review the substantial literature on the subject.
Three insights which emerge from the process of reflection are:
1. The specific position which will be arrived at is that a positive research and teaching link primarily depends on the nature of students’ learning experiences, resulting from appropriate teaching and learning processes, rather than on particular inputs or outcomes.
2. In certain circumstances, a positive link may well depend strongly on the abilities of students.
3. A rational explanation of the persistent myth of a general positive link will be given.
4. All good teaching is an act of translation – from the level of the teacher to that of the learner. And as in all translating, there is inevitably some distortion.
All these insights depend of course on the assumption that teachers wish to link their teaching to research; teachers – and there are many who legitimately hold this view – who see their primary role in developing their students’ abilities in eg employment oriented and life skills inevitably have other priorities.
Mythology and Early Studies
There is a strongly held belief among the majority of academics that teaching in higher education should take place in a research atmosphere (see eg Jensen 1988, Millar 1991, Smeby 1998), although the opposite is rarely stressed. While this belief is vocal, it is only as a belief that it is substantiated by research (see eg Neumann 1992, 1993, Robertson 1999) and can therefore justifiably be called a mythology. More worryingly, the myth has become - at least in Britain - institutionalised through the finding of an apparent positive relationship between research and teaching, based on the external evaluations of teaching and research in British universities. This finding which, if confirmed, could have serious consequences on university funding, can however be explained much more readily and convincingly on the grounds that (a) assessors tended to consider research performance ipso facto evidence of good teaching, a matter that deserves further investigation, (b) leading research universities had for many years been better funded and had substantially more favourable staff-student ratios (HEFCE 1995, paras 94 - 112), and (c) the inspection teams were predominantly academics who were respected by their peers in the main for their professionalism in research which was public knowledge; their teaching abilities being largely unknown in an academic culture where teaching is an essentially private activity conducted by amateurs.This latter point is mentioned largely to show that enquiries into the problem of the research-teaching link have been bedevilled by simplistic investigations which would not pass muster in any other research area. Unfortunately, the current interest in the relationship between teaching and research has been biassed by managerial considerations. If there is no positive link, then the two can be functionally separated; universities can divide their staff into researchers and teachers and no doubt pay the latter less. This in turn reinforces the myth in the eyes of academics who may fear for their jobs. Not a good climate in which to conduct dispassionate enquiry.
Two other deeply flawed forms of investigations should be mentioned. The first consists of enquiries from practising academics (see eg Jensen 1988) which in reality investigate in the main their personal views, often expressed with considerable strength, and thus reinforce the mythology of the subject; the second consists of investigations in the so-called ‘scientific’ mode, in which uni-dimensional quantitative performance indicators for individuals - such as the number of publications for research and scores on student evaluation forms for teaching - are correlated. Unsurprisingly, results proved contradictory (Flood Page 1972, 1985, Ramsden and Moses 1992). More recently, starting with Ben-David (1977), this extraordinarily naive approach has been replaced by more sophisticated and mainly qualitative approaches. Excellent summaries of what is known from detailed research, both quantitative and qualitative, regarding the research and teaching link, have been provided by Hattie and Marsh (1996), Neumann (1996) and Jenkins (1999).
The Programme for the University of Berlin
The starting point for the present analysis of the link, which - as has been stated - will not produce new research data, will be what may well be the oldest relevant statement on the subject. It occurs in the monograph of Humboldt on the future of the new University of Berlin. Humboldt (1810) wrote - and the passage is quoted first in German, as it loses in translation:
“Es ist ferner eine Eigentümlichkeit der höheren wissenschaftlichen Anstalten, daß sie die Wissenschaft immer als ein noch nicht ganz aufgelöstes Problem behandeln und daher immer im Forschen bleiben, da die Schule es nur mit fertigen und abgemachten Kenntnissen zu tun hat und lernt. Das Verhältnis zwischen Lehrer und Schüler wird daher durchaus ein anderes als vorher. Der erstere ist nicht für die letzteren, beide sind für die Wissenschaft da.”
In my translation of this very densely packed passage, in which I to some extent follow an earlier translation (Humboldt 1970), I will use both ‘scholarship’ and ‘learning’ as being closest in different ways and in different circumstances to the difficult word ‘Wissenschaft’, where ‘scholarship’ is defined as a deep understanding of a subject and ‘learning’ is both a process and an outcome - for both teachers and students - of the application of scholarship to research or teaching (for an excellent analysis of the relationship between German academic words and possible English equivalents, see Pritchard 1998):
“It is furthermore a peculiarity of the institutions of higher learning that they treat higher learning always in terms of not yet completely solved problems, remaining at all times in a research mode [ie being engaged in an unceasing process of inquiry]. Schools, in contrast, treat only closed and settled bodies of knowledge. The relationship between teacher and learner is therefore completely different in higher learning from what it is in schools. At the higher level, the teacher is not there for the sake of the student, both have their justification in the service of scholarship.”
In one stroke, a master stroke, Humboldt has abolished the problematic nature of the research-teaching link per definitionem: university teaching only deserves that title if it involves a joint endeavour between teacher and learner, between professor and student, in a common search for knowledge. The outcome of both research and teaching is new learning, wholly new in the case of research and new to the learners - who may traditionally be labelled ‘professors’ and/or ‘students’ - in the case of teaching. In spite of what Humboldt says, some of this undoubtedly also goes on now in schools, but it is in the nature of schools and of those in them that a school can never become as totally a community of learning as universities can and should. In schools predominantly and rightly, teachers teach and students learn. Whether the same should be the case in the earlier undergraduate years is highly arguable; Humboldt would certainly not have thought so.
This could be the end of this paper, but it turns out that there is after all more to the link than was dreamt of in Humboldt’s philosophy.
Multi-dimensionality and measures
The first step in all good practice of problem solving consists of further analysis and perhaps redefinition of the problem. In the process it will be appropriate to descend from Humboldt’s Olympian heights and recognise the realities of universities at the turn of the millennium.
Firstly, it is now accepted that the relationship between research and teaching is multi-dimensional. Some of the dimensions are indicated by the following questions:
•Whether the unit of assessment is an individual, a department or an institution (Moses undated; Elton 1998, on teaching; McNay 1999, on research);
•Whether the comparison is at the level of competence or excellence? (Elton 1996a, on teaching; HEFCE1996, on research);
•Whether the comparison is being made from the perspectives of academic staff (see eg Millar et al 1991), students (Jenkins et al 1998), administrators (Neumann 1993) or funding bodies (McNay 1999);
•Whether the comparison depends on the discipline investigated and if so, how (Moses 1990);
•Whether there are cultural factors which are different in different institutions (Jenkins et al 1998) and different countries (Gellert, Leitner and Schramm 1990, Whiston and Geiger 1992);
•Whether different performance measures yield different results (Tomlin 1998, Ham & Marshall 1998).
Only some of these points can be addressed within the limits of this paper.
Secondly, possible measures have to be identified, presumably not simplistically quantitative ones, on which a comparison can be made. How can one estimate whether either research or teaching is in some sense ‘good’?
What is considered ‘good’ will undoubtedly depend on who judges it so. In the case of the research-teaching link, there are at least three very different kinds of stake holders who can take on the role of judges:
•academic staff, who believe that their professional judgment enables them to identify what is ‘good’. This has been characterised as ‘connoisseurship’ (Eisner 1985);
•students, whose satisfaction with what is offered to them depends on their motivation which can be very varied (Elton 1988, 1996b, Neumann 1994, Jenkins et al 1998);
•administrators and funding bodies, who - in different ways - are preoccupied with considerations of cost effectiveness (Neumann 1992, 1993).
Clearly, the research-teaching link may be seen quite differently by different stake holders. Research into it has so far been based almost wholly on the perceptions of academic staff, and these, ever since the seminal work of Halsey and Trow (1971), have been consistent and, incidentally, essentially independent of age. The views of other stake holders are only just beginning to be considered (see eg Brown 1998) and, while I am happy to leave to others a discussion of the legitimate interests of administrators and funding bodies, much more will be said below about the interests of students.
Student interests are of three kinds:
1. The self-interest in passing examinations, which teachers must never devalue. It is one in which students naturally take the lead.
2. Interest in the subject, which is primarily a task for teachers to develop.
3. Interest in developing abilities which will be of value long after leaving formal education. These abilities may be ‘academic’, eg the ability to analyse, synthesise and make judgments (see eg Ramsden 1992, p. 20 for an extended list), or related to life skills, eg the ability to see a task through to its conclusion (see eg Association of Graduate Recruiters 1995, p. 21 for an extended list). The development of such abilities requires in general a close collaboration between teachers and students.
All three kinds of interest are wholly legitimate; all too often it is the third that is neglected.
Correlations and interpretations
As has been stated, the bulk of the early research on the teaching-research link has been quantitative, but such work can at best establish relationships, not causality. There are four ways in which such relationships may be connected causally:
•Good research causes good teaching
•Good teaching causes good research
•The relationship between good research and good teaching is dialectic, ie they support each other
•There are one or more additional factors which, if present, cause the correlation.
To distinguish between these four possibilities, it is helpful to postulate an underlying theory. Since the research has produced such varied and even contradictory results, it would appear most natural to search for a theory which is based on the last of these possibilities and to look for a factor or factors, which may or may not be present in a particular investigation.
Two such factors have indeed been proposed, without which a relationship may not be expected. The first of these, which goes back to Humboldt, is ‘scholarship’, ie a deep understanding of what is currently known in a discipline and which illuminates both research and teaching in that discipline (Elton 1986). The second is ‘learning’, ie good research and good teaching both lead to good learning: in the first case of something that is genuinely new, in the second case of something that is new to the learner (Brew and Boud 1995). Clearly, the first of these, scholarship, is an input, ie it is argued that without the input of scholarship good learning does not occur. The second, learning, is not only an output, but, as Brew and Boud rightly stress, also a process. Thus it may be postulated that an input of scholarship is necessary, if a correlation between ‘good’ research and ‘good’ teaching is to exist, and that if it does exist, then it can be verified through the process of learning and through the ‘good’ learning that follows. [Recognition that the word ‘learning’ can be both a verb and a noun, is important here.]
It should be noted that while learning as a process is accepted in conjunction with teaching, this is not so in research where, as Brew and Boud point out, research is normally presented as the communication of ideas detached from the processes from which they are derived. Their conclusion is that “teaching and research are correlated where they are co-related, ie when what is being related are two aspects of the same activity:learning”.
Ben David (1977) concluded that the teaching-research relationship can be both supportive and conflicting, and that the balance of these is strongly subject dependent. In particular, he contrasted the humanities, where originality lay mainly in creative scholarship, with the sciences, where it lay in the discovery of new knowledge through experimental research, and suggested that it is only in the former that a strong positive link between research and teaching was and could be established. In contrast, Schwartzman (1984) argued that in the Germanic - but perhaps not always in the Anglo-Saxon - context, the pursuit of both science and the humanities includes a component, namely scholarship (or Wissenschaft), which links teaching and research. This conclusion, in which of course Schwartzman follows Humboldt, although he does not quote him, led to the heretical suggestion (Elton 1986) that scholarship is perhaps insufficiently valued in the Anglo-Saxon pursuit of science.
However, scholarship should be viewed (Boyer 1987, Rice 1991, Elton 1992) as not only supporting research (the creation of new knowledge) and teaching (the transmission of knowledge [2]), but also practice (the application of knowledge) and integration (the synthesis of knowledge). All academic activities, including incidentally the management of knowledge, should be influenced by scholarship and, if they are, then it may well be true that there exists this link between them and research (in its widest sense). This extension of the meaning of scholarship, and also of research, is a generalisation, in the sense that mathematicians use this word, which raises the question as to whether it allows the extension of a possible research and teaching link to these generalised concepts. There is little if any research on this so far.
Having identified scholarship as the input, learning (verb) as the process and learning (noun) as the output, there remains the question of the more detailed nature of the process which links input to output. What forms of teaching, what forms of learning, what curricula in the widest sense, can lead from scholarship to learning?Now, there is a very basic point of learning theory, namely that learning with understanding, so-called ‘deep’ learning, requires learners to integrate new knowledge with existing knowledge (for a discussion of the evidence see eg Ramsden 1992, ch. 4). For this to happen, students must be actively involved in the learning process and indeed come - at least in part - to own it. This is the very antithesis of learning through didactic teaching, which in most instances can lead to only superficial learning. Unfortunately, much of assessment confines itself to assessing such superficial learning and hence a change in assessment methods forms an important aspect of any desired new curriculum. The conclusion that the nature of the link between research and teaching depends primarily on the process of the student curriculum rather than on the outcomes in either research or teaching is arguably the most important insight obtained in this paper.