1. If universities are to widen participation and provide access to more students of all ages, they will need to find pedagogically acceptable combinations of teaching methods and cost structures to sustain increased student numbers. As long as Government subsidises teaching costs, there is an upper limit on the amount of subsidy that can be available to HE at any one time. Because of this upper limit, expansion creates a downward pressure on the teaching cost per student. In the recent rapid expansion this has resulted in a shift from teaching methods with high variable costs to those with lower variable costs - usually from small group teaching to large group teaching, from individual supervision to group supervision. Each new student represents a high marginal cost to the institution for methods where the staff-student ratio is high, so these methods have been cut, with a resulting inevitable loss of quality of learning experience for students.

2. Universities have also been experimenting with new technology, in the expectation that this could lead to more cost-effective teaching. Almost invariably IT has led to higher costs, with greater efficiency still a promise for the future. One solution is to find an alternative cost curve for teaching methods that allows institutions to expand numbers without loss of quality of the learning experience, and with improved efficiency. Academic staff time is the principal cost of teaching in HE, and the analysis below looks at the effect of the different methods on staff costs, as student numbers increase.

The analysis

3. To simplify the analysis, we consider how students' learning time might be distributed across different combinations of three teaching methods: lectures, small groups and resource-based learning (RBL). RBL must be differentiated into two types, given their very different costs: 'external' refers to materials produced external to the university, such as textbooks, commercial CDs and software; 'in-house' refers to materials developed in-house, such as printed lecture notes, courseware programs, customised spreadsheets.

4. The model used here makes the following assumptions about staff time needed for one hour of student learning time:

Lectures 3 hours preparation + 1 hour presentation = 4 hours
Groups 1/2 hour preparation + 1 hour presentation = 1.5 hours
RBL (external) 2 hours preparation for use of existing resources
RBL (in-house) 20 hours for developing in-house.

5. This model assumes a maximum of 100 students per lecture, and 10 students per small group. Resource-based learning preparation and development requires the same amount of time, regardless of the size of the student group. 2 hours preparation time would be appropriate for selecting materials for students to read and work through, with a written study guide. Twenty hours would be appropriate for developing materials with low production costs. (Many IT-based developments would take 1-200 hours per student study hour.) The question is: 'How do the cost structures of these methods behave with respect to staff time needed as student numbers increase?'

6. As student numbers increase, the cost per student falls, so that the total cost stays within public spending constraints. Chart 1 models what has happened in many institutions. For example, for a course with 100 hours of teaching time, a ratio of 30:70 lectures:groups, (traditionally a preferred combination for many types of course), doubling the student numbers considerably increases costs, whereas a 70:30 combination produces a much lower increase in total cost, due to the lower variable cost of lectures. Similarly, courses with a high proportion of small groups, and low-cost use of resources, such as library work (70:30 groups:RBL) are also on a cost curve that virtually doubles the total cost when student numbers double, even starting from a lower base. Institutions have consequently moved to a lower cost combination of more large group teaching, even though it offers less to students.

7. An alternative solution for dealing with increased student numbers has been the creation of more resource-based learning materials. They offer more interactivity than books, and make use of interactive programmes and communications technologies. With negligible variable costs, the relatively high production costs of these methods can be amortised over large numbers of students to achieve a better cost curve, capable of withstanding expansion. These resources do not exist, and for most staff attempting to exploit these methods, the preparation time is greater than for traditional resources as it includes development time. Chart 2 shows that developing RBL (RDev) does put the course on a different cost curve, so that increasing student numbers will make little difference to the total cost (the production costs are 20 hrs: 1 hr). No institution would attempt that amount of development, but the graph serves to illustrate where institutions are positioning themselves if they move towards RBL developed in-house. A very different picture emerges from adopting existing materials where staff time is costed only for preparation (RUse), and the cost curve begins low and retains the shallow gradient for larger student numbers.

8. From the above analysis it seems that the most efficient combination of teaching methods sustainable for increasing student numbers, will fulfil the following criteria:

  • improve rather than diminish students' learning experience;
  • increase the use of fixed cost rather than variable cost methods;
  • increase RBL developed externally rather than in-house.

9. To take a more complex and realistic combination of methods, the following table outlines a 'Traditional' model for a course, with lectures, a high proportion of small group work, use of resources such as library work, and a small amount of in-house resource development. Table 1 gives a likely distribution of student time across those methods for the kind of university course for which the system was originally designed.

10. Pressure of numbers has resulted in many courses moving to a structure closer to the 'Current' model in Table 1: small group work has given way to large groups; library/resource centre use is unchanged; but lecturers are also trying to introduce more in-house development of teaching materials as a way of coping with numbers, and exploiting new media. One clear disadvantage with this model is that the amount of small group work has dropped. A possible 'Future' model is also outlined, which restores group work to a more acceptable level, retains some lectures, but also relies heavily on use of external RBL, while preserving a small amount of in-house development.

11. Clearly, movement in the direction of the 'Current' model tends to diminish rather than enhance the student's learning experience, because of the reduced amount of group work. Preservation of the 'Traditional' model, however, leads to a worsening cost curve as numbers increase (see Chart 3). The 'Current' model will have been successful in keeping total costs within funding constraints, as can be seen from the flatter gradient in Chart 3, but the price paid in terms of diminished student learning experience makes this unacceptable as a model for the future. The alternative outlined in Table 1 achieves both a flatter cost curve than the 'Traditional' model and a lower overall cost, ie the total cost for 100 students on the 'Future' course is similar to the total cost for 50 students on the 'Traditional' course.

12. This simplified analysis can only be indicative. A full analysis for a particular course will take into account a wider range of staff costs eg teaching time, assessment time, project supervision, fieldwork, lab-work, personal guidance, etc. This is the kind of modelling the HE sector will have to undertake if expansion is to be achieved within reasonable cost.

13. It is clear from the above that continued expansion will require a radical shift in the way university teaching is carried out. The implications of the 'Future' model are far-reaching. To achieve that kind of cost curve, while preserving staff-student contact, it was necessary to increase RBL to cover the majority of student learning time. For that to be acceptable, the resource materials must be very well-designed, (whether print, audio visual, or software), and capable of supporting a student working independently. Traditional text-book style will not suffice.

14. The new interactive media, offering adaptive feedback and student control have the potential to support independent study, but only if fully developed, tested and maintained. This kind of development requires a staff:student time ratio of more than 100:1, rather than the 20:1 assumed above. Many staff would seek to spend some of their time on development of learning materials, because these will enshrine the core of their teaching. But for the high quality materials needed for the majority of university study, development will have to be of a high order, and therefore not in-house, in general. IT methods must achieve their promise of greater efficiency both by improving the quality of student learning, and by amortising the cost of development over large student numbers.

The National Committee would like to thank the following contributors to this Appendix: Professor David Chiddick, Professor Diana Laurillard, Sir George Quigley and Mr David Wolf.