Paper to Nordisk konferens - Forskning om distansutbilding, Nuläge och visioner. Umeå, June 14th - 16th, 1993
My contribution should primarily be regarded as a case description on why and how one specific institution, NKI, has applied systematic research related to practice and development work. The main aim has been to obtain information and build up a knowledge base for quality development. Research into different aspects of distance education has been a continuous activity in NKI for 20 years, based on the shared belief in the organisation that confirmed knowledge represents one of the necessary prerequisites for systematic work to increase the quality of teaching.
NKI is today the second largest non-governmental educational institution in Norway including two state accredited colleges (engineering and computer science), a unit for decentralised face-to-face vocational courses, a publishing house and NKI Distance Education.
Although NKI could be called a dual mode institution, the Distance Education Unit is operated as a large scale single mode distance teaching institution. My paper concerns research related to distance education only, and as most of our work has arisen out of problems and challenges related to large scale teaching, I have chosen to point this out in the title. The main message is, however, that I believe that the policy of continuously integrating formal research and evaluation with the normal development and operation of distance teaching carry large benefits in any type of institution.
Large front end investments in course development, large target groups - and relatively smaller costs involved when actually carrying out the teaching process are typical for large scale institutions.
NKI research has to a large extent been related to questions around completion and drop out, as we have defined increased completion rates as the main criterion of success. This focus might be discussed, and it has met some criticism.
It was only natural that the first focus of interest was surveys on recruitment and completion, as systematic statistics at that time were lacking both nationally and internationally. As a result of the information gained, the focus changed to research on student support and follow up systems, technology related projects and evaluation studies (Figure 1).
Figure 1. Research activities at NKI (Not available in the web-version)
The focus of the research over time has partly been a result of changes in the needs of the institution and partly a result of general developments in the field of distance education and priorities of fund granting agencies, such as the Ministry of Education, Research and Church Affairs. I should also be noted that our research to a large extent has taken place in co-operation with other institutions and organisations, such as our friendly competitor, NKS Distance Education, and Telecom Research in Norway. NKI and NKS also took the initiative to establish the Norwegian Centre for Distance Education (SEFU) aiming at stimulating co-operation on research and distribution of information on distance education. The Norwegian State Institution for Distance Education (NFU) has later joined SEFU. Thus, NKI and NKS has carried out a number of research projects in co-operation. Examples are research on tutor feedback, computer mediated communication, satellite distribution, and evaluation of video conferencing, telephone and fax applications.
Most of the research carried out at NKI may be characterised as evaluation studies. At the same time we have regularly carried out routine evaluation of specific courses and/or study material. Thus, in the overview in Figure 1, we have illustrated this by the dotted line. During the last few years, NKI has carried out some larger evaluation studies applying more systematic evaluation research methodology (Peersen 1992, Lakskjønn 1993).
In general, NKI research has been driven by practice rather than theory. This means that our studies has aimed at solving practical problems, rather than testing theories. This priority seems to be accepted, see e.g. Moore 1985:
"Distance education is an applied professional field, and such fields - like medicine - need two kinds of research. Obviously it needs research which helps solve problems, but it also needs basic research - that is, research which tests and extends basic knowledge which is helpful to us all - and that is the theory." (Moore 1985, p. 41)
In many ways, large scale distance education is suited as a "laboratory" for experimental testing of different methods, with clearly formulated hypotheses and well defined variables, randomly selected groups and with measurable results on study success, drop out and completions. Moore (1985) argues for experimental research in distance education:
"If we want to know the effect of a particular teaching technique, the researcher must control the effects of other conditions and manipulate the technique in question and see the effect on the learners. This is the only sure way of evaluating practice, yet it is very seriously neglected in distance education." (Moore 1985, p.42)
On the other hand experimental research has certain drawbacks. Many problems just do not lend themselves to experimental investigation. Secondly, when carrying out an experiment, one cannot make many changes without ruining the design and the possibilities of drawing conclusions. Often, it might be of more interest to examine the process, and to be able to respond to the process with changes and improvements. The experiment may interfere with rational and desirable changes for the students involved. Thus, in many instances other than experimental methods should be preferred. This has also been the case in the recent NKI studies of technology and new media were pilot try outs, developmental testing of new systems (such as computer mediated communication) and evaluations with more open ended problems and questions have initiated the projects. All experiments at NKI have been followed up by questionnaires to survey opinions and attitudes to the different treatments.
Although the NKI research has been practice driven, it has generally been founded in implicit theories based on practical experiences and common views on teaching and learning. Many studies have been designed as experiments to test relevant hypotheses, and as such may have contributed to build up theory or to put theories to test. As I read Holmberg (1991), the NKI experiments on "turn-around time" and "the personal tutor/counsellor" are mentioned as a support of hypotheses derived from the theory of "guided didactic conversation" (Holmberg 1983), or supporting a model of distance education based on student guidance or "continuity of concern" (Sewart 1980).
Similarly, Bajtelsmit (1988) refers to NKI surveys and experimental studies when developing his "conceptual model for attrition in distance education. Another recent example is Peters' (1992) reference to the "personal tutor/counsellor experiment" when suggesting to test the drop out models of Bajtelsmit and Tinto (1975) at the Fernuniversität.
We have mentioned that drop out has been a focal point of our research. On some occasions we have been criticised for being too occupied with drop out in distance education. We believe that we are in agreement with most distance educators that reducing drop out is a major challenge in our field (se e.g. Peters 1992). This fact must not be taken as a support of the view that drop out is a larger problem in distance education than in other types of part time education. As far as we know there is no evidence supporting such an assumption.
For the institution drop out may be a considerable financial problem. Through economic analyses Keegan (1990) indicates that the viability of an educational institution depends very directly on the number of drop outs in the system.
A McKinsey report on an American institution ..."focused on student attrition as a deficit-producing trend that threatened the very future of this distance institution." (Bajtelsmit 1988, p.3).
From an individual student's point of view, Bajtelsmit holds the position that "the negative effects of dropout are obvious: loss of opportunity for personal and career advancement, lowered self- esteem, and increased likelihood of future disengagement." (Ibid. p. 2.)
During the preceding years research on drop out in education has largely applied a model often referred to as "Tinto's (1975, 1987) theory".
The theory explains the persistence/withdrawal process which depends on how well the student becomes involved in the social and academic processes of the academic institution. The model describes the concepts and four sets of variables in a causal sequence:
Figure 2. Tinto's (1975) model for drop out from college (Not available in the web-version)
The student enters the academic institution with a social and personal background that influences which commitments he/she will have to the institution and to complete the studies. These background characteristics and initial commitments will influence how the student will perform and get involved in the academic and social systems. The experiences of academic and social nature during the studies will interact with the background variables and subsequently influence the student's later academic and goal commitments. According to Tinto it is the students integration into the social and academic systems of the institution that most directly relates to continuance/withdrawal.
The model has mainly been applied in research on attrition in full time education, but it has also been referred to and/or applied in studies on distance education (see e. g. Bååth 1981, Sweet 1986, Taylor et al. 1986).
Bajtelsmit (1988) has questioned whether Tinto's theoretical model is appropriate for use with non-traditional students, such as part time distance students. He proposes a model for explaining and predicting drop out in distance education that puts more emphasis on the influence of the external environment, specifically the student's occupation and family, while the concept of social integration in the institution is given a less prominent role. Bajtelsmit does not devalue the importance of academic support in the distance study setting, but shifts the primary focus ..."from the socialization process of previous models to the congruencies and compensatory relationships between the educational (academic) and external (occupational) subsystems." (Ibid. p. 13.)
Figure 3. Bajtelsmit's (1988) model of drop out from distance education (Not available in the web-version)
Without going into details of this model, we may note that it emphasises individual background and "independent/distance learning skills" and academic support systems as major variable group on the institutional side. These are similar to the variables that we have concentrated on in NKI research to identify prospective drop out students and to increase completion rates, and we believe that Bajtelsmit's model represents a better framework for analysing drop out and completion in distance education.
NKI survey studies (Rekkedal 1972a, 1973, 1976, 1978) have examined questions concerning recruitment, completion and attrition, comparative studies of distance and full time study in similar areas of education, and relations between background variables (i.e. variables grouped by Tinto (1975) and Bajtelsmit (1988) as background variables and initial commitments), and study success.
The results from these studies may very shortly be summarized in the following statements:
There is also a positive correlation between level of previous education and all measures of success in distance study, and a negative relationship between time since last school experience and success.
Thus, specific measures should be taken to support younger students, students with low previous education and students who have been away from school for some time.
As mentioned NKI has put some emphasis on carrying out experimental research testing different hypotheses on teaching techniques, support and guidance of distance students.
The experiments carried out at NKI have generally been carried out according to the following design (in some instances more than one experimental group have been included):
R : E X1 O1
R : C X2 O2
R means that the students have been selected to the groups at random
E is experimental group
C is control group
X1 stands for the variable that is examined, while
X2 stands for the treatment given to the control group
O1 and O2 stands for measures of results, such as achievements/study success/completions
In the following we shall give some examples of experimental studies carried out at NKI:
Turn-around time is defined as the time elapsed between the student mailing the assignment till it is received by the student corrected and commented upon by the tutor. In the experiment, the turn-around time was reduced in an experimental group relative to a control group, without the tutor knowing who belonged to which group. According to information from the students on a questionnaire, the experienced turn-around time was reduced from a median of 8.3 days to 5.6 days. The result was a measured increase in completion rates from 69% among the control students to 91% among the experimental students, the difference being statistical significant at .001-level. According to student responses, the critical limit of turn-around time to be accepted as satisfying was one week, a conclusion which seems to be supported by later research (Rekkedal 1983). This experiment was chosen as basis for an international co-operative project organised by the ICDE Research Committee aiming to replicate previous research in different settings (Taylor et al. 1986).
In another experiment we tried to examine the effect of following up inactive students by a sequence of postcards and motivational letters. The sequence was started automatically when a student had not submitted assignments for one calendar month, and stopped when the student started to submit assignments again - or established contact with a student advisor. Differences in study activity between 240 students in each of a randomly selected control and an experimental group were measured. We found clear and statistical significant differences between the two groups after three months. During the third month of the experiment, 46% of the experimental and 31% of the control group students were active studying and/or in contact with their advisor (Rekkedal 1972b).
The above project initiated another experiment involving 3 experimental groups and a control group to examine the effect of introducing an introductory course in study techniques and systematic follow-up of new students. This experiment concluded carefully that the course in study techniques combined with initial follow-up may reduce the number early drop-outs (Rekkedal & Hallem 1975). The experimental studies also include a study on the use of pre-produced tutor comments and standard solutions in distance education, as well as a couple of media research studies, "the telephone as a medium for tutoring and guidance" (Rekkedal 1989) and "telefax for two-way communication" (Rekkedal 1992).
It seems clear that the experiments have led to changes in the regular operation of the distance teaching system at NKI. In some cases the differences measured have been clear and significant, and it has been concluded that the experimental variable really has had an effect on study results - and student attitudes. In some cases, however, an administrator might also take decisions based on non-significant differences in similar ways as he/she would respond to other types of information available.
This experiment arose out of experiences from the above mentioned projects, literature search and a specific aim of serving new students in a better way.
During the planning stage we carried out some intensive group interviews with several new students. These interviews showed that the students seemed to be generally satisfied with their experiences in distance study.
The students reported, however, one common difficulty: They were reluctant to contact the administration, the counsellors or their tutors when they met problems, and they were uncertain about whom to contact in order to seek advice on different problems.
We felt that the rationalisation and "industrialisation" of distance education that seem to be necessary to cater for large student groups at acceptable costs in a "large scale" distance teaching system, probably result in a division of labour and a depersonalisation of the teaching processes. This situation seemed to cause larger difficulties for many students than we were aware of.
In the early days of correspondence education, the schools were often established by enthusiastic persons stimulated by an idea or a specific situation. What these people lacked in formal pedagogical knowledge, they compensated by entrepeneurial vision and devotion to their work. As professional distance teaching organisations developed, some of what we earned in theoretical orientation and efficient systems might have been lost in reduced personal teaching/learning relationships.
Thus, we decided to design an experiment where we wished to personalise teaching by introducing what we called "the personal tutor/counsellor".
In the experiment one person integrated the roles of different tutors in different subjects, as well as student advisors/counsellors. Within the responsibilities of this tutor we also tried to include other measures which we believed were important to help the distance learner to complete his/her studies successfully. In short, we tried to construct a system which would constitute a "new" way of organising the tutor's work during the initial phases of the study period, the main emphasis put on increasing the experienced quality of the didactic functions of the distance tutor and the two-way communication between the tutor and the student. In total, 10 different aspects of tutor-student interaction were included in the experimental variable.
The students selected for the experiment were assigned to a personal tutor who followed them closely during the 3 to 11 first courses of the study programme for a certain qualification. By this formal change in organisation, a number of aspects of the tutor's work and the division of labour between the administration, the counsellors and the tutor were changed.
We also selected a control group, that was not deprived of any of the services which had been introduced on the basis of theory and research during the preceding years, and which at the time constituted a normal part of the NKI distance teaching system. The main difference in the treatment of the experimental and the control group was that the experimental students were taken care of by one personal tutor integrating administrative, teaching and counselling functions, which normally were divided between different persons, departments and specialists.
Data were collected from the normal NKI files and study records, and minute records kept by the tutor on the communication with the students, follow-up letters, special measures, telephone calls etc. A small questionnaire was developed to assess the students' attitudes. It contained questions on correspondence study in general, the tutors' work and counselling to examine whether different aspects of being a correspondence student where experienced differently by the students in the two groups.
The students were selected into the experiment during a period of 5 months. Data on study progress were collected at the time of completion of the first defined stage of the study programme, at the time of cancellation or exactly 8 months after enrolment (for students who neither had completed or dropped out).
It is easy to see that the experiment has some drawbacks concerning control of sources of error, which variable that cause possible effects, and generalisations to other systems. We were clearly aware of these problems when deciding to carry out the project. We simply wanted to examine a total system, instead of looking at isolated variables - the aim being to see if the new organisation "could produce a better total quality system". Consequently, we find that possible generalisations to other systems must origin in the basic ideas, rather than in specific findings.
The experimental role of the tutor is described below. The experiment covered 10 different aspects of the tutors work.
| Aspect | Experimental group | Control group |
| 1. Tutor | Same tutor during the first 3-11 courses | Different tutors in different courses |
| 2. Employment | Permantly employed full office time | Part-time employment at home, paid per assignment |
| 3. Tutoring/counselling | Same person responsible for all student communication | Responsible for written assignments only, other staff for general counselling |
| 4. Turn-around time | Assignments returned the same day from the school | Assignments sent via the tutor's home address |
| 5. Study technique | Same tutor teaches study techniques | Specific part-time tutor in study techniques |
| 6. Follow-up of new students | Tutor takes contact with all new students via mail or phone | Automatic routines with form letters |
| 7. General follow-up | Tutor takes contact with all inactive students via mail or phone | Automatic sequence of form letters |
| 8. Telephone tutoring | Students may phone the tutor. Tutor calls when needed | No systematic use of telephone tutoring |
| 9. Tutor presentation | Personal presentation with photo and phone numbers enclosed with the study material | Presentation of each tutor enclosed with the first assignment returned from the tutor in each separate course |
| 10. Pre-produced tutor comments | Deleoped for all courses. Applied when needed | May have been used by som tutors |
Figure 4. Aspects included in the "personal tutor/counsellor experiment"
As a result of the experiment we hoped to find some answers to the following questions:
Will this organisation have any effects on the number of students starting their studies (defined by submitting one or more assignments)? Will this organisation have any effects on drop-out rates during the early stages of study? Will this organisation have any effects on the students' pace of study and general study activity? Will this organisation affect the student' attitudes toward correspondence study in general and/or on specific aspects of correspondence study? What will the consequences of this new tutor role be regarding the organisation of other aspects of the distance education system as well as the operational costs of the system?
We found marked and significant differences between the groups. 8 months after enrolment the experimental group had a significant higher rate of completion. The number of active students were comparable in the two groups, while the control group had a significantly higher number of inactive students. Similar results were found after one year of study. The experimental students were more active in their studies and had completed a larger number of study units and single courses during the experimental period.
Concerning "start/non-start" we found no significant differences between the groups. The non-start rate was "low" in both groups, 9 and 13 percent, respectively. Neither, did we find any significant difference concerning lapse of time between enrolment and registration of the first assignment.
In general the experimental students reported more positive attitudes towards "correspondence study in general", "the study material" and "the work on assignments", aspects not included in the project. None of these differences were, however, statistically significant. This was hardly unexpected, as these aspects were not part of the experimental variables. Concerning "feelings of isolation" we found less problems in the experimental group, but again the differences were not statistically significant. This might be surprising as one could expect that the more personal communication that we hoped to create by the integrated personal tutor/counsellor role would reduce possible feelings of isolation in the studies.
On the other hand we found clear and significant differences on the following areas:
The experimental group students expressed more positive attitudes towards "the general quality of the tutor's work", "assistance and support from the tutor", "guidance and counselling form the institute, tutor or counsellor" and "follow-up by the institute and/or the tutor", and whether "telephone tutoring is of any help in distance study".
The work on the "personal tutor/counsellor" concept and this experiment resulted in NKI organising the initial phases of study according to these principles. We found that introducing the personal tutor/counsellor on a normal basis was both financially and pedagogically sound. However, during the last couple of years organisational changes, new media and some difficulties concerning cost and efficiency control of the tutor/counsellor role in NKI have gradually lead to different organisational, administrative and teaching structures. Basic ideas from the project has survived and are implemented by larger emphasis on the counselling functions of the distance tutor, whether part-time or full-time, and in connection with the introduction of new media and methods, such as systematic telephone tutoring, telefax communication, computer mediated communication and two-way video conferencing.
From the early eighties NKI has carried out research and try outs of different kinds of technology, to some extent in co-operation with other Norwegian institutions, such as NKS and the Norwegian Telecom Research. This research has covered technologies such as telephone and telefax applications (Rekkedal 1989, 1992), video, television and two way video conferencing (Holden 1992) and computer mediated communication (Paulsen 1989a, 1989b, Paulsen & Rekkedal 1990a, 1990b, Bjørgen 1992).
The EKKO Project on computer mediated communication in distance education may stand as one example of testing of new media and communication technology. In spite of quite modest expectations of revolutionary changes when introducing new media in the distance learning system, we looked upon computer communication and conferencing as specifically promising. We found that many of the ideas of personal communication examined and supported in previous experiments, such as "the personal tutor/counsellor", could be built into a system of computer mediated communication, and we believed that the medium could be introduced within the frames of a large scale system without too many difficulties.
The NKI EKKO Project (the acronym EKKO stands for "electronic combined education" in Norwegian) started in 1986. The aim of the project was to: develop a computer based conferencing system for distance learning and experiment with it in different contexts to gain pedagogical and administrative/organisational experiences within distance education based on computer conferencing - in order to install conferencing as a standard option for NKI distance students.
The project has been organised as a long term development task following these stages:
The EKKO Project has covered system development and theory studies as well as many small surveys. The project has produced some research reports and many articles and conference papers. One specific student thesis analysed problems concerning adaptation and integration of distance education technology in the student's family. Morten Flate Paulsen, initiator of the EKKO Project, is presently working on a doctoral thesis on "pedagogical techniques for computer-mediated communication" (Paulsen 1993).
After more than 5 years of research into computer conferencing, it is now offered as a standard option in some NKI courses. Through our research we have learned that computer mediated communication as a technology is quite different from other media in distance education. However, we have not drawn the conclusion that computer conferencing in itself represents a new generation of distance education (Nipper 1989).
We have experienced that there are large differences in introducing computer conferencing in large scale and small scale distance education systems. Distance teaching by computer conferencing puts large demands on tutor time and efforts. Increased costs in the teaching phase should ideally be compensated by reduced costs involved in course development. We have found that it is difficult to reduce development costs without reducing the quality. Thus, computer conferencing in a large scale system may easily become very expensive.
To be able to exploit the possibilities of group activities we have decided to organise the computer conferencing courses with fixed starting dates and fixed progression schedules, which has led to computer conferencing being less flexible than correspondence education. It seems that many students prefer the flexibility and individual freedom of correspondence study, while others stress the advantage of group communication in computer conferencing as a necessary condition for their participation in distance education. We see great challenges in trying to develop computer conferencing to take care of both the social needs of some students and the needs of other students who prefer the flexibility of individual study with possibilities of more efficient communication.
In this paper we have mentioned some of the NKI research projects on distance education. Some have been reported in some more detail that others. Pure evaluation studies have only been mentioned.
Finally, we shall try to summarise some of our general experiences and point to some important challenges for the near future.
Through our research at NKI we definitely have learned that research and evaluation as a continuous activity related to the practice of developing and carrying out distance education is an important means in quality development.
The research staff is continuously influenced by ideas coming from the teaching and administrative staff. And from the other side - the activities involved in research and evaluation stimulate the development of new ideas and interests and an openness among the teaching and administrative staff to change their practice.
There is one important point which I would like to stress. Systematic collection of data and reporting, in our experience, is the absolute fundamental basis for achieving lasting changes and quality developments. We have seen that potential good ideas may be tried out and lead to positive change of practice. However, if one does not take the time or the resources to really examine and report the effect of specific changes and developments by methods which people trust, changes seem to have a tendency to be very short-lived. When people in the system are actively involved in the developments, and the developments are proven to be effective by accepted methods of evaluation, those developments tend to be followed up - and live.
Before pointing out some challenges for the future, let us take a look at a total model of distance education, giving an overview of different media, methods and communication technologies:
Figure 5. Models, methods, media and communication forms in distance education
The following overview lists the media and technologies available, (not necessarily complete as new technologies or modifications are introduced continuously).
| Presentation media Printed material Correspondence by mail Audio cassette Radio Television CD-rom |
Media for two way communication Telephone Audio conferencing Video conferencing Video phone Audio cassette Audio graphics Electronic mail |
| Computer communication Computer conferencing |
Self instructional material Learning packages/Learning programmes Computer based learning programmes Interactive video Laboratory "kits"/"trainers" CD-rom |
| Direct teaching Classroom teaching Seminars Laboratory training |
Group learning Study groups without tutor |
Figure 6. Media, methods and technology in distance education.
Institutions active in distance teaching are in some ways under continuous pressure, mainly from the technology and communication industry, to adopt new technologies into existing distribution systems. The large challenge for large scale distance teaching institutions is to gain experience in how the possibilities new media and new technologies can be implemented to increase the quality of the learning experience, without having negative effects on proven teaching quality and economics of the large scale system. These institutions have administrative systems, routines, and economy which are difficult to adapt to some of the modern technologies. We must realise that the development of pedagogic and didactics adapted to new technologies seems to take much more time than the technological developments, and that, in fact, we know very little about how different distance students actually learn and how different media influence the learning process. At the same time, even within the educational sector, it seems to be much more difficult to fund pedagogical driven than technological driven projects. The real large challenge is to try to change this situation.
The question of future research needs was discussed at a Norwegian conference on 'Research programme for distance education' (SEFU 1989) and also at a Nordic seminar one and a half year ago, where Bååth (1991) suggested some important areas for future research. Some of the areas in need for research mentioned on these occasions were:
In our view, all these areas are in need for research with specific reference to Nordic policies, practices, educational traditions, and the new media and communication technologies which are introduced.
It should be stressed that future research in distance education in the Nordic countries should build upon previous research and development work reported from our region and internationally. There is a danger of waste of time and money when new groups enter the field and start experimenting without taking the effort to examine the experiences that already have been collected. It is also of interest to 'replicate' studies done in other contexts to see whether experiences gained under different circumstances are valid when further developing new distance and flexible learning systems in the Nordic countries.
Bajtelsmit, J. W. (1988): Predicting Distance Learning Dropouts: Testing a Conceptual Model of Attrition in Distance Education. Report to the ICDE Research Committee.
Bjørgen, A. M. (1992): Teknologi og hverdagsliv. Fjernstudenters bruk og opplevelse av datakonferansesystemet EKKO med familien som læringsmiljø. Bekkestua: NKI/SEFU
Bååth, J. A. (1981): A comment on Rekkedal's report: 'Introducing the personal tutor/counsellor in the system of distance education. Epistolodidaktika, 1, pp. 36-38.
Harris, W. J. A. (1975): The Distance Tutor. Education by correspondence. Manchester Monographs 10. University of Manchester.
Holden, G. (1992): Videokonferanser som del av undervisningsopplegg i fjernundervisning. Bekkestua: NKI/SEFU.
Holmberg, B. (1983): Guided didactic conversation in distance education. In Sewart, D. et al. (eds.): Distance Education. International Perspectives. London: Croom Helm.
Holmberg, B. (1991): The feasibility of a predictive theory of distance education: What are we allowed to expect. In Holmberg, B. & Ortner, G. E. (eds.): Research into distance education. Frankfurt: Peter Lang.
Holmberg, B. & Ortner, G. E. (eds.): Research into distance education. Frankfurt: Peter Lang.
Keegan, D. (1990): The Foundations of Distance Education. London: Routledge.
Lakskjønn, K. H.: Evaluering av Bedriftsøkonomistudiet ved NKI Fjernundervisningen. Bekkestua: NKI.
Moore, M. (1985): Some observations on current research in distance education. Epistolodidaktika, 1, 35-51.
Nipper, S. (1989): Third generatiuon distance learning and computer conferencing. I Mason, R. & Kaye, A. (1989): Mindweave. Communications, Computers and Distance Education. London: Pergamon.
Peersen, V. S. (1992): Evaluering av høgskolestudiet "Ledelse og administrasjon for helse- og sosialsektoren. Bekkestua: NKI/SEFU.
Paulsen, M. F. (1989a): En virtuell skole. Del I. Fundamentet i EKKO-prosjektet. Bekkestua: NKI.
Paulsen, M. F. (1989b): En virtuell skole. Del II. Erfaringer fra EKKO-prosjektet. Bekkestua: NKI.
Paulsen, M. F. (1993): Pedagogical Techniques for Computer-mediated Communication. Pedagogical Online Seminar, Oslo, February 12, 1993.
Paulsen, M. F. & Rekkedal, T. (1990a): The Electronic College. Selected articles from the EKKO project. Bekkestua: NKI.
Paulsen, M. F. & Rekkedal, T. (1990b): Den elektroniske høgskolen. EKKO-prosjektet, Del III. Bekkestua: NKI.
Peters, O. (1992): Some observations on dropping out in distance education. Distance Ed., 13, 2, 234-269.
Raynor, T. E. (1985): Identification of potential distance learning drop-outs. Report to the ICDE Research Committee.
Rekkedal, T. (1972a): Correspondence studies. Recruitment, achievements and discontinuation. Epistolodidaktika, 2, 3-38.
Rekkedal, T. (1972b): Systematisk elevoppfølging. En eksperimentell undersøkelse av virkningen av kontaktbrev til elever ved NKI-skolen. Oslo: NKI. (English summary in Epistolodidaktika, 1973, 2, pp. 57-63.)
Rekkedal, T. (1973): Three years' enrolments to NKI-skolen's correspondence courses. Recruitment and discontinuation trends. English summary in "Tre årskull brevskoleelever. Stabekk: NKI.
Rekkedal, T. (1976): Tekniske studier. Korrespondanseundervisning og klasseundervisning. Delrapport 1: Rekruttering. Stabekk: NKI.
Rekkedal, T. (1978): Tekniske studier. Korrespondanseundervisning og klasseundervisning. Delrapport 2: Brevskoleelevene. Stabekk: NKI.
Rekkedal, T. (1983): The written assignments in correspondence education. Effects of reducing turn-around time. An experimental study. Distance Education, 4, 2, pp. 231-252.
Rekkedal, T. (1985): Introducing the personal tutor/counsellor in the system of distance education. Project report 2: Final report. Stabekk: NKI.
Rekkedal, T. (1989): The Telephone as a Medium for Instruction and Guidance in Distance Education. Bekkestua: NKI/SEFU.
Rekkedal, T. (1992): Telefax in individual distance study. English summary in Telefax som medium for toveis kommunikasjon i individuell fjernundervisning. Bekkestua: NKI/SEFU.
Rekkedal, T. & Hallem, S. A. (1975): Follow-up of correspondence students. The effect of an introductory course in study technique and encouraging letters to new enrollees. English summary in Begynneroppfølging av brevskoleelever. Oslo: NKI.
Sewart, D. (1980): Creating an information base for an individualized support system in distance education. Distance Ed., 1, 2, 171-187.
Sewart, D., Keegan, D. & Holmberg, B. (eds.): Distance Education. International Perspectives. London: Croom Helm.
Sweet, R. (1986): Student dropout in distance education: An application of Tinto's model. Distance Education, 7, 2, pp. 201-213.
Taylor, J. C. et al. (1986) Student persistence in distance education: A cross-cultural, multi-institutional perspective. Distance Education, 7, 1, pp. 68-91.
Tinto, V. (1975): Dropout from higher education: A theoretical synthesis of recent research. Review of Educational Research, 45, 89-125.
Tinto, V. (1986): Leaving College: Rethinking the causes and cures of student attrition. Chicago: University of Chicago Press.