‘Rescuing the personal tutor: lessons in costs and benefits’
Ormond Simpson
Chapter in ‘Perspectives on Personal Tutoring in Mass Higher Education’(2006)
ed. Thomas and Hixenbaugh. Pub Trentham Books ISBN-10:1-85856-385-2
Rescuing the personal tutor: lessons in costs and benefits
Ormond Simpson
Introduction
“Follow the money” – ‘Deep Throat’
There are many excellent reasons for promoting the idea of personal tutoring in UK higher education. Some arguments centre around essentially ethical ideas of giving students a fair and just deal, some arguments around the fact that as participation widens increasing numbers of less well-prepared students may enter higher education and require increased support and so on. This chapter will attempt to argue the case for personal tutoring from a financial perspective – in the words of Bernstein and Woodward’s secret source in the Watergate affair ‘Deep Throat’, it will attempt to “Follow the money”.
This approach is important because senior university budget holders are more likely to be influenced by arguments that invoke financial logic. This is not to be cynical about university management:any budget holder faced with conflicting demands must consider carefully where investment is most effective and personal tutoring is an expensive activity. If there is an argument that personal tutoring not only makes practical sense but can be justified financially then it is much more likely to attract senior management support. This chapter will attempt to show that personal tutoring not only makes sense financially to institutions and students, but that there are circumstances in which it can even return a financial surplus to the institution.
The first step in ‘following the money’is to try to link personal tutoring with student retention.
Personal tutoring and student retention.
Student retention is a complex affair in which there are many variables. Trying to separate out the effects of a personal tutoring system is very difficult. Perhaps the only way would be to undertake a‘medical’trial in which two groups of students of similar characteristics were compared, one group receiving no personaltutoring (or a placebo, although it’s difficult to imagine what form that might take), the other receiving some kind of personal tutoring. Such a trial would hardly be ethical. Another way might be to introduce a personal tutoring scheme into a student support system which had hitherto not been organised that way and see if there were subsequent differences in student performance. Such ‘experiments’ would still be difficult given the many variables that make up student success. But inadvertently the Open University set up the conditions for such a trial when it changed its student support systems in around 1996, in effect abolishing its previous personal tutoring system.
The personal tutor in the Open University
When setting up the Open University in the late 60’s its planning group recognised that that the many low qualified students entering the university would probably require high levels of support. They therefore designed a personal tutor system as part of its support programme. These personal tutors were called ‘counsellors’ and their remit was to give ‘non-academic support’ (Simpson, 2002) to students, helping them adapt to distance learning and progress their studies. Counsellors were therefore very much the equivalent of the ‘personal tutor’ in conventional higher education.
In addition OU students were allocated a course tutor for their particular course who was responsible for teaching that course. When students moved onto their next course they were allocated another tutor but the counsellor remained responsible for their non-academic support (OU students may take 6 to 12 courses over 4 to 8 years for their degree). Tutors and counsellors were both part-time members of the OU’s staff.
From the late 70’s the roles of tutor and counsellor were amalgamated into a ‘tutor-counsellor’ role, where the tutor appointed for a new student’s first course was also their counsellor and remained so for the rest of the student’s studies. This system was thought to work well; tutor-counsellors were able to make good relationships with their students whilst teaching them during their first year, and those relationships often endured for the remainder of the student's career.
However there were few attempts to collect evidence of the value of the tutor-counsellor role. Simpson (1977) found that students overwhelmingly agreed with the statement ‘there should be a person in the OU system who has a personal knowledge of their progress and to whom students could refer.’ But this finding could be criticised on the grounds that this was a difficult statement to disagree with and that no alternatives were offered. Thus when change to the role was proposed there was little objective evidence of its value.
The effective elimination of the personal tutor in the OU
Proposals for change arose because of changes in OU entry. Prior to 1992 nearly all students entered on the same range of courses – the OU’s five Foundation courses – and tutor counsellors were therefore appointed to these courses. When in the early 90’s the university started allowing students to enter on any one of its 150 courses the picture became more complex. There were two possible options for changing the counselling system:
- students entering on non-foundation courses could be allocated a counsellor who wouldn’t be their tutor. This was already the case for ‘associate students’ studying one-off courses and it was clear that the relationship built up between counsellor and student was less satisfactory given the lack of a teaching contact;
- all tutors could be designated tutor-counsellors. But apart from the increased payment costs involved it was thought that not all the OU’s 5000 tutors would want to take on this link.
It was also believed that as the OU’s programmes of study became more complex that it was unreasonable to expect part-time staff to be able to answer the whole range of questions that students might ask. However it was also clear that the counselling system was thought by senior management to be too expensive - although no figures were ever published.
So from 1996 over a period of years the counselling role of the tutor-counsellor was eliminated and counselling was taken over by full time staff in the OU’s 13 regional centres on a reactive basis, with teams of three or four academic-related staff supported by clerical staff servicing the needs of anything up to 15,000 students. Students still had a different course tutor for each course they took and could refer to full time staff by phone, email and letter for non-academic support – an essentially ‘call centre’ model of student support with no personal counselling.
To complete the picture the University recently invested a large sum – rumoured to be of the order of several million pounds - in a ‘Customer Relation Management (CRM) system. This software maintains student records and allows the University to see what contacts the student has made with the University and what topics have been raised.
At no stage in any of these changes was any effort made to evaluate their effects on student learning. Such evaluations would have been difficult since the multi-variate nature of the causes of student dropout is well-known (Woodley, 1987), but given the sums of money involved and the self-styled character of the university as a ‘learning organisation’ it is surprising that no attempts were made to see if the modifications had been effective. Perhaps in the absence of any clear theory of student support an institution's policies are likely to be the result of the most powerful voices in the system.
Nevertheless there were dissenting voices to the changes, albeit from student-facing staff. In general such staff are usually denied the knowledge – particularly the financial knowledge - to make cases against developments sponsored by senior management. And once senior management have backed a project then a version of the ‘Titanic Effect’ (Watt, 1974) may come into operation – no-one looks out for icebergs.
Eliminating the personal tutor – the consequences.
Despite the lack of evaluation it may now have become possible to draw some conclusions from the abolition of the personal counsellor:
- Student retention on course. Student retention rates in the UKOU have been drifting downwards for a number of years but the rate at which they were declining appeared to accelerate somewhat over the period at which the tutor-counsellor role was eliminated – from around 1996 onwards – see figure 1 which uses data from the OU’s Technology courses as an example.
Fig. 1 Student Retention (%) in the UKOU on successive
Technology foundation courses
- Student re-registration. In 2003 nearly 30% of new students completing their first year did not carry on to a second: in fact the proportion of students carrying on after their first year has been dropping since the early nineties - see figure 2.
Fig. 2 Re-registration rates (%) of new students completing the previous year
Once again there is a small but definite drop in re-registration rates from 1996 onwards (although admittedly small in comparison with the long term drop from 1990).
Thus in two measures of student retention there are small but clear drops after 1996. These cannot be sourced definitely to the abolition of the personal counsellor but there were no other significant changes in the university at that time so at the very least that possibilty must be kept in view.
The ‘personal tutor’ makes a come-back
Considerations like these led to attempts by junior staff to evaluate the effect of re-introducing the personal tutor in some way in small scale projects. One of the first to report was Simpson (2004). Around 3000 new students were identified as being particularly vulnerable to dropout by a statistical analysis of their entry characteristics. Half were proactively phoned before their courses started by a ‘study adviser’. The call was relatively short (an average of 10 minutes) and its content aimed at making a relationship with the student and enhancing both their integration with the institution and their motivation using findings from the relatively recent field of ‘Positive Psychology’(Snyder et al, 2001). The other half with the same levels of vulnerability was kept as a control group.
The experiment was repeated in three successive years (Table 1).
Year / Contacted group total / Contacted group withdrawn during course / % of contacted group withdrawn / Control group total / Control group with drawn during course / % of control group withdrawn / Difference in withdrawal rates contacted - control2002 / 1433 / 272 / 18.98 / 1433 / 328 / 22.89 / 3.91
2003 / 1152 / 184 / 15.97 / 204 / 43 / 21.08 / 5.11
2004 / 910 / 158 / 17.36 / 97 / 21 / 21.64 / 4.28
TOTALS / 3495 / 614 / 17.57 / 1734 / 392 / 22.61 / 5.04
Table 1 Differences in pass rates of contacted and non-contacted students 2002-2004
It can be seen that on average the experimental group had a retention rate around 5% higher then the control group. This may not sound like a large increase in retention. But students admitted to the OU are likely to be older than conventional students, have lower qualifications on entry, to be in work and have family commitments. They are therefore more likely to experience institutionally unavoidable dropout due to illness, job changes, family commitments and so on. In fact it has been estimated that a 5% increase in retention is nearly a third of the maximum possible increase in retention that is possible through institutional activities although this estimate is open to question (Simpson, 2003).
It may strain credulity to suggest that this effect was due to one phone call. But the effects are clear and appear to be continuing into 2005. There are reasons however why such a simple contact could have had such an effect:
- Student dropout in the OU is heavily front-loaded. About 15% of students dropout before course start and another 30% in the period before submitting the first assignment. They therefore have little contact with their tutor and any teaching activity. Thus to have much effect on retention it is likely that an activity be focused before or at course start - as this one was.
- There is no personal proactive contact from the university before course start. Potential students are encouraged to enrol with the university on the Web and are therefore unlikely to have any personal contact with the university before course materials arrive. Thus a personal interactive contact stands out and may consequently have a marked result.
- The style of the contact was anindividually focused activity aimed at enhancing the student’s motivation rather than on identifying weaknesses. A contact based in to some extent on evidence derived from anappropriate evidential basis in psychology may again have more effect than contacts without such a basis.
Costs and benefits of the OU study adviser activity to the institution
Since this exercise was carried out in isolation with balanced control groups it is not unreasonable to infer a direct causal effect between the contact and the retention increase. In consequence it is possible to isolate the costs and to a lesser extent the financial benefit of the contact and calculate its return on investment. This can be done as follows:
Consider an activity applied to N students which costs £P per student and produces an increase in retention of r%. The total cost of the activity will be £NP and the number of extra students retained will be Nr/100. The ‘cost per student retained’ is then £NP/(Nr/100) which equals 100P/r.
In the OU exercise the increase in retention as indicated previously was around 5%. The cost of the individual contact was estimated to be about £10 in staff time and overheads (average call length was of the order of 10 minutes and staff made around 3 calls an hour, the remainder of the time being used for recording and administration). Thus the ‘cost per student retained’ was 100x10/5 = £200.
At first sight this figure is intimidating. It is hard to imagine a university budget holder allowing such expenditure especially if applied to all new students. In the OU with 35,000 new students a year the overall expenditure would be £35,000x10 = £350,000 for an increase in retention of 35,000x5/100 = 1650 students which does not sound like a bargain.
But as well as costs it is important to estimate benefits. In common with other UK universities the OU receives a grant from the Higher Education Funding Council for England based on the number of students who complete their courses (the funding regime is more complex than this statement implies but it is sufficiently accurate for this analysis). This sum is currently about £1100 for a student completing a 60 credit point course. In addition there may be savings on recruitment costs. If dropout is reduced then fewer students need to be recruited to maintain steady numbers. Currently it costs the OU around £500 per new student recruited in marketing costs. If it assumed that a proportion of that – say £200 - might be saved through increased retention then the total income to the OU for every student retained is around £1300.
Thus the ‘surplus’ income for every student retained is £(1300 – 200) = £1100 which is a return on investment of 1100/200 = 550%. If the activity was to be applied to all new students and a similar level of retention achieved for a similar cost then the total surplus income to the university would be £1650x1100 = £1.8m. Clearly this figure is very approximate but nevertheless it does appear that it is possible to increase retention at least at zero cost and possibly even make some kind of small surplus.
The attraction of not only increasing student retention but possibly making a surplus was the point at which OU senior university management started to take notice and consequently this activity is being mainstreamed throughout the university from 2005 on. Clearly further proactive contact beyond the initial call might increase the retention further and some limited work is being undertaken to investigate this possibility. But with more contacts there mayeventually be an effect of diminishing returns – Case and Elliot (1997) suggest from an evaluation of an online programme in the US that the optimum number of contacts is five but how far this finding is reliable or applicable to conventional UK higher education remains to be seen.
As well as mainstreaming this activity the OU is to try to use the same adviser to make the proactive contact with the same student each time. An obvious next step after that will be to make the study adviser reactively contactable by the student. If and when those characteristics are established the OU will have in effect reintroduced its personal tutor system some years after it was formally abolished.
It is unwise to draw substantial conclusions about personal tutoring in UK universities from one activity in a non-typical university such as the OU. There is clearly a very large difference between a system that is based on one proactive phone call and full personal tutor system. Nevertheless there is an implication, to put it no more strongly, that, as Chapman (2003) says “to increase retention you must personalise your relationship with your students” and that an increase in retention can go a long way to funding such personalisation.
Costs and benefits of personal tutoring to students
Clearly it is not only the institution which has a financial interest in student retention. From 2006 students will be investing substantially larger sums in their education through increased fees as well as (in full time education) suffering loss of earnings over that study period. It is notoriously difficult to estimate the return on that investment but various authorities have attempted to assess the ‘graduate premium‘ – the amount extra a graduate could hope to earn over a lifetime compared with a non-graduate. Figures from Walker and Zhu (2003) for example suggest a graduate premium of £200,000. If the Return on Investment to students is defined as the graduate premium divided by their initial investment then that RoI is around 600%. It is widely thought that this RoI must decrease with higher levels of participation in higher education but evidence (Simpson, 2006) from countries such as Sweden and Australia which already have higher levels suggest that the RoI is holding up reasonably steadily. (In the OU the RoI for student investment appears to be higher than for conventional universities at around 2200%. This is due mostly to the absence of the loss of earnings factor although the lower fees also have an effect).