Survey

As individuals interested in doctoral education and committed to academic development, we wanted to ensure that whatever project we developed would be embedded in its context. The institutional visits and the literature review both helped us to ensure a rigorous and engaged deliberation about our doctoral practices and to help us to develop what we believe to be a very strong proposal. To this we added a survey undertaken as Rhodes University and University of Venda in which we invited postgraduate supervisors and scholars to anonymously answer a few key questions.

We developed the survey collaboratively after an initial Skype planning meeting and we piloted the questionnaire within the team. Chrissie Boughey then circulated the questionnaire to academics at Rhodes University, Sioux McKenna circulated the relevant link to PG students at Rhodes University and Langutani Masehela circulated the links to both supervisors and students at University of Venda.

The remainder of this page presents a few of the findings from the survey.

Neither institution is close to the goal of 75% of staff having doctorates by 2030 (NDP 2011). While the two institutions have similar numbers of postgraduates, they have very different proportions of postgraduates within the student body. Rhodes University's 30% postgraduates is part of the reason it is designated 'research intensive' in much of the literature (Cloete, Mouton and Sheppard 2016). It is worth noting that despite having almost double the number of student that Rhodes University has, University of Venda has nowhere close to double the number of permanent academic staff numbers. The rise of casualisation of academic staff is another matter of concern with 64% of academic staff on temporary contracts in South Africa (CHE 2018).

The sampling for this survey was simply by widespread notice to the university communities. The findings need to be treated with caution given the numbers of not necessarily representative of the whole staff or PG student body. We believe the data raises a number of useful issues which we have taken into account in the planning of the project.

While the sample size prohibits broad generalisations, it is significant that none of the Rhodes University supervisors did not have a doctorate. Academics are generally expected to have a PhD at the point of employment or would be expected to already be working towards one. This has implications for the selection of candidates for this project as we may be less likely to have academics at Rhodes University who are not already registered for a PhD. It is anticipated that most applicants will come from outside of the research-intensive institutions.

It is clear that the crude 80:20 rule is at play in both institutions whereby the bulk of the supervision is undertaken by a few people. In each case, four or five people had successfully graduated a large number of individuals. It is also clear that while there is a spread of experience in supervising at Master's level, this is less the case at doctoral level. 88% of the University of Venda staff and 42% of the Rhodes University supervisors have not supervised a single doctorate to completion.

However, as can be seen in the table below, the supervision load is fairly high (albeit vastly uneven with a spread from 0 to 18). Only 21% of the University of Venda respondents (7 supervisors) and 26% of the Rhodes University respondents (16 supervisors) are not currently supervising at doctoral level.

The two tables above illustrate how the respondents indicated they had been inducted into their role as supervisor and then what support is available at their universities for supervisors. It is important to note that 47 Rhodes University supervisors indicated awareness of the Strengthening Postgraduate Supervision course but only 4 had actually completed it. They cited lack of time despite interest and positive evaluations from their colleagues or postdoctoral fellows and PhD candidates who had completed it. Such support structures need to be built into the workload (or project plan) if they are to be widely used.

The role of mentoring in supervision capacity development was most evident through both the closed and open ended responses. In most cases this referred to informal mentoring (sometimes in a co-supervision structure). The absence of induction was noted by a large number of staff at Rhodes University, alongside fears regarding personal capacity. The project needs to use a clear combination of formal support and mentoring to assist novice supervisors.

There was much mention in the Rhodes University data of the CPGS workshops and short courses for postgraduate scholars, and a number of mentions of the Research design course. It would seem that this new academic development initiative is beginning to gain traction. Three people indicated that these offerings were too generic and not sufficiently slanted towards the Natural Sciences - an issue to bear in mind in future programming. There was also great support for the Higher Degrees Guide which outlines all processes and roles and responsibilities.

There were a number of comments from the University of Venda participants that there was not enough support for postgraduate students and supervisors generally. Wherever possible, initiatives developed in University of Venda within the project need to be opened to the whole university community.

Supervisors often utilise various models of postgraduate supervision for different students at any one time. Despite some respondents indicating that they used project teams and so on, the vast majority relied on one-on-one supervision, co-supervision or one-on-one supervision with some support seminars. These models all have the potential to exacerbate power differentials and to lead to feelings of isolation for the student. They also provide very little support for the supervisor built into the model. See the Literature Review for more discussion on this. Those supervisors who did use project team approaches came from the Natural Sciences, the Business School and Education at Rhodes University and from research centres at the University of Venda.

While many similar issues emerged across the two institutions in response to the open ended question about challenges experienced by supervisors, there were also some notable differences. While the quality off students was mentioned by academics at Rhodes University, this was far less so than at University of Venda where it was very noticeably the most common issue. Students were generally pathologised in regards language proficiency, research skills, and lack of motivation. One respondent noted that international students were far stronger than South African students.

Funding was the biggest concern for Rhodes University supervisors, perhaps because postgraduate students have traditionally been on campus and fully funded. More research is needed on this.

Rhodes University and University of Venda supervisors both mentioned workload issues as a major constraint. Balancing various academic roles alongside the need to develop one's own research was seen to be particularly onerous. In Humanities and Social Sciences where supervision is often on topics unrelated to the supervisor's own research focus was seen to be problematic. Developing project teams for doctoral education is one means of developing one's research profile alongside one's supervision.

Interestingly, eleven Rhodes University supervisors mentioned issues of feedback. They noted the time and expertise required to give the kind of feedback that made epistemological access possible. They expressed concerns about their role in developing students as academic writers. The giving of feedback and the development of student writing is an area noted as especially key to strong supervision in the Strengthening Postgraduate Supervision course where it is given particular attention. Both students and staff in the proposed project will be able to benefit from completing this course.

Bureaucratic issues regarding increased administrative demands and inefficient systems was mentioned by a number of academics at University of Venda and with particular reference to ethical clearance processes at Rhodes University.

The responses to this question indicated an awareness that being an active researcher is fundamental to being a good researcher. Positioning the postgraduate student as a novice member of the field was also seen to be key.

The postgraduate student survey was completed by 93 students at Rhodes University and 19 students at University of Venda.

While 18% of the postgraduate students at Rhodes University are also staff members, 74% of the postgraduate students at University of Venda are staff members. This relates to the drive to get academics to get doctorates, the lower levels of qualifications of current staff and the differing employment requirements.

As indicated earlier, the supervisors at both institutions indicated that 80% of their supervision is through one-on-one supervision, with or without a co-supervisor, and as indicated above, more than 90% of the students at both institutions are currently being supervised through one-on-one supervision, with or without a co-supervisor. This survey included those in the Natural Sciences where working in teams in the laboratory or field is common and so this percentage is even more significant than if the survey was only completed in the Humanities and Social Sciences. Indeed, looking at the raw data indicates that if we exclude those in the Natural Sciences, it is very close to 100% of students being supervised in this way. Given the concerns raised in the South African literature that the dominance of this model plays a role in low throughput and retention rates (for example ASSAf 2010, CHE/CREST 2009, Cloete, Mouton & Sheppard 2016, Backhouse 2010), this needs careful consideration. It is hoped that our work with Lancaster University to develop more structured and supportive models of doctoral education can prove an important space for change.

How was your supervisor allocated or selected?

At Rhodes University, the vast majority (87 out of 93) approached the supervisor of their choice, and at University of Venda, 14 out of 19 did the same. They indicated that they approached their choice of supervisor because they:

•Had worked with them previously

•Knew their work or searched out an expert in the field

•Applied for a particular project (often funded)

In six cases at Rhodes University (out of 93), the student submitted a general application and concept note/proposal to the institution and was then allocated a supervisor. Five cases (out of 19) were similarly allocated at University of Venda.

It is clear that selection of supervisor is usually the prerogative of the student. Bearing in mind the plans for our project and the ways in which supervision allocation is handled at Lancaster University, this issue will need to be carefully managed.

How helpful are support structures?

Rhodes University students were generally aware of policies, HDG, courses and workshops available. A few students (17) indicated they could not attend due to time pressures or because they are off-campus. Three students (out of 93) indicated the workshops were too generic for them to attend. The majority indicated that the workshops were very useful to:

•Gain confidence, try out ideas

•Network with other PG scholars

•Learn how to write and how to use certain software for data analysis

Most University of Venda students (12 out of 19) were aware of policies and seminars and the remainder indicated there was not support available.

Greatest Challenges

Data across the two institutions was very similar.

Most frequently mentioned issue was funding. Trying to cover fees and living expenses, even issues like inter-library loans, difficulties of being an international student and costs involved.

Time was also a very frequently mentioned issue. Balancing multiple jobs, family responsibilities, and so on.

Mental health issues raised by a number of students: imposter syndrome, loneliness, isolation (especially but not only off-campus students), anxiety, depression, lack of confidence.

Poor supervision mentioned by a few students: waiting too long for feedback, conflict between supervisors, getting little feedback, supervisor being unavailable (always away on conferences).

Workload expected of the students in terms of reading and writing mentioned as too onerous by four students.

Ethical clearance processes seen as a particular hurdle.

Constraining institutional culture and lack of transformation mentioned by one student.

Greatest Strengths?

Data from the students across the two institutions was very similar. They listed:

• Resilience and patience, motivation and commitment.
• Work ethic.
• Perseverance.
• Knowledge of the field.
• Writing abilities.

It is hoped that working in a structured, supportive environment as a member of a larger team might ameliorate some of the challenges listed above. The isolation of the 'lonely scholar' model of doctoral education (Backhouse 2010, Harrison 2010) has been implicated in many of the mental health issues experienced by doctoral candidates. The collegial support for tackling issues of institutional culture and processes such as ethical clearance should also prove a strength of the proposed model. The explicit focus on culture (the valuing of respect, collegiality, support) alongside the project structures (the project team approach, the use of coursework, the Doc Weeks and online seminars etc) should allow us to celebrate and make the most of the strengths identified by the students who completed the survey.