Part 3

Lessons of Experience

 3.9 Who bears the cost of learner support? 

In all of the programmes reviewed for this research, learner support is seen as an integral part of the service and therefore budgeted as part the institution’s costs in providing the programme. Learners are not expected to pay extra for support services as such, although telephone contact tends to be learner initiated and therefore at the learners’ expense, and additional costs are incurred by learners in transport to and, sometimes accommodation at, contact sessions, as well as the cost of posting in assignments if no courier service is available. 

The ACCESSDL&T: ECD programme is of particular relevance to a discussion of costs, since it will close down at the end of 1999 as the cost of the operation cannot be met by the fees payable by learners.  

The ACCESSDL&T programme is in many ways quite unique in so far as it involves a private provider investing in a wide range of support services, rather than relying on the correspondence model favoured by most private providers.  

Although numbers of learners at some of its decentralised centres is quite small in some cases, the fact that the programme can no longer be financially sustained may have more to do with the nature of the programme and its target audience than anything else. An ECD programme aimed at such a wide range of interested parties inevitably requires a wide variety of support mechanisms to meet the diverse needs of learners on the programme, but it is targeted at a user group who are unlikely to be able to afford the fees needed to cover the cost of maintaining such a wide range of support services. An ECD programme is a far cry from an MBA programme, which is currently the growth area for privately-offered ODL courses in South Africa, where course fees average at about R45 000 regardless of whether one studies full or part-time. It is worth noting at this point that the UNISA: ABET programme requires a one-third external funding subsidy to bridge the gap between the cost of running the programme and the fees that learners are able to pay. It is also worth noting that at any one time on the WITS: FDE programme a significant number of learners are behind with their fees payments. Financial management of a distance-learning programme is therefore a critical area of competence. 

Since research has so far failed to make a direct link between the provision of different kinds of leaner support in terms of which form(s) of support are most cost-effective, it is very difficult to simply say that ACCESSDL&T should simply have had fewer centres used by more learners and should have offered a less extensive range of support services. It may well be that more learners would then have dropped out of the programme as arrangements would no longer have been so accommodating and this could actually result in an increase in the per capita cost of producing a successful learner.  

There is clearly a need for more research in regard to the relationship between learner support and programme budgeting. South Africa has many years’ experience of distance education provision and several institutions, including some of the largest in the country, are beginning to move away from the correspondence model and to build in learner support structures, of which direct human interaction is often a key component. South Africa therefore is ideally placed at this time to undertake some longitudinal studies on the links between provision of learner support and the per capita cost of successful programme graduates.  

Common sense, and some of the limited empirical data available, suggests that the up-front cost of setting up learner support mechanisms, and quality assurance monitoring of those structures, will be recouped in the long term through increased retention rates, higher success rates and improved quality of learning. However, the economic laws of diminishing returns also indicate that there must be an optimum level of support beyond which increased funding for support is likely to begin to increase the per capita costs of success.

Finding the optimum level of investment in this area is complicated further by the fact that the optimum support system will vary from programme to programme, from culture to culture and from target audience to target audience. 

In order to offer some food for thought for thought in this debate, provided below is an analysis drawn up by Glennie and Butcher of SAIDE, using a spreadsheet model developed by Butcher, on some budgetary implications of changes in the elements that make up a theoretical learner support system.  

While by no means a definitive model at present, and based upon a number of assumptions that we do not have the space to go into here, some of the implications are nevertheless illuminating. 

Costing courses

SAIDE’s course costing tool allows one to determine the costs of a course on the basis of putting values to a very wide range of variables.

These include:

·     learner numbers

·     duration (in weeks)

·     number of times the course is offered in one year

·     notional learning hours

·     variety of ‘contact’ strategies: lectures, tutorials, practicals, individual consultations, etc

·     peer group study

·     independent study

·     varieties of assessment strategies: short hand long assignments, journals, examinations, etc all prepared, marked and moderated

·     course design time per hour of study time

·     number of years for which design remains valid

·     research time

·     instructional design and editorial time

·     production costs including audio tape, video tapes, printed texts, photocopied readers, internet material etc

·     technology used

·     sundry

·     overheads. 

In what follows, we have identified four basic scenarios and variants on these. Across the four basic scenarios we have kept as many variables as possible constant, (e.g. 200 notional learning hours, examination strategy etc). The variants within the scenarios only alter learner numbers. 

In order to move from our course (200 notional learning hours) costs to full time equivalent year costs (1200 notional learning hours), we have simply multiplied by six and added certain other costs such as programme co-ordination. We are currently working on a database that will assist us to develop programme costs more easily. 

We have not included the full 20-page spreadsheet for each scenario and variant. We have simply provided the consolidated page for each course and for each full time equivalent year. The primary variables are described in the summary table which follows immediately below. 

SUMMARY TABLE

Scenario 1: Distance Education

Course Design:   50 hours of course design time per study hour, 172 pages of designed course material, 1 x 120 minutes, video, one textbook; course revised every three year.

Support: 8 hours of tutoring per learner per course in groups of 25; 1 hour of individual tutoring per cours

Assessment: 1short assignment and1long assignment, both marked by tutor and moderated by educator; 1 examination marked by educator and externally moderated.

Costs

		Number of students	Cost per 200 n.l.hr course	Cost per f.t.e
Scenario 1		2000	R 1664	R 10115
Scenario 1A		500	R 2882	R 17445
Scenario 1B		100	R 9385	R 56562

Scenario 2: Correspondence

Course design: as in Scenario 1

Support: none

Assessment: as in Scenario 1

Costs

		Number of students	Cost per 200 n.l.hr course	Cost per f.t.e
Scenario 2		2000	R 1134	R 6931
Scenario 2A		500	R 2353	R 14720

 Scenario 3: Correspondence with low level of assessment

Course design: in Scenario 1

Support: none

Assessment: examination only

Costs

		Number of students	Cost per 200 n.l.hr course	Cost per f.t.e
Scenario 3		2000	R 822	R 5061

 Scenario 4: Primarily face to face

Course design: 10 hours course design per study hour

                        100pp of course pack material

                        course reviewed every three years

Assessment: as in Scenario 1

Support:  32 hrs of tutorials per course in groups of 25

Costs

		Number of students	Cost per 200 n.l.hr course	Cost per f.t.e
Scenario 4		100	R 4259	R 25689
Scenario 4A		50	R 6627	R 39910
Scenario 4B		500	R 2409	R 14578

Some comments by way of initial analysis

• Scenario 1 and its variants show how extensively distance education course costs vary with enrolment. It would be interesting to compare the actual per capita costs on the fledgling UOFS: BML programme with 48 learners, the UNISA: ABET programme with about 5 000 learners and the ACCESSDL&T programme, and to explore how the UOFS: BML could be amortised over time.

• Scenario 2 shows that taking away the substantial learner support included in Scenario 1 reduces costs considerably (by 20% in the case of 500 learners, and by 30% in the case of 2000 learners). It is this cost differential that currently distinguishes private providers, of which ACCESSDL&T is a notable exception, and state-funded providers.

• Scenario 3 shows that even if you spend quite considerable time on course design, straight correspondence with limited assessment is very cheap. However, all of the programmes reviewed have rejected a straightforward correspondence model, and their provision of a more expensive comprehensive and varied learner support structure seems to have paid dividends in that they all enjoy very high retention and success rates, even though all the programmes are subject to some form of external evaluation and quality control.

• Across Scenarios 1, 2 and 4, we can see that costs for 500 students are in the same ‘ball park’. As numbers go up distance education as described in Scenario 1 becomes cheaper. As numbers go down to say 100, distance education as described in Scenario 1 becomes much more expensive than face-to-face as described in scenario 4. Clearly, good distance learning, which implies the provision of adequate learner support, is not a cheap option. However, high up front costs can be amortised over time as learner numbers increase.