Silence Is Golden: Using ‘Safe Words’ to Promote Research Student Ownership in Supervisory Meetings

Andrew James Clements, School of Psychology, University of Bedfordshire Gail Kinman, School of Psychology, University of Bedfordshire


The quality of supervisory relationships has a significant impact on research students’ ability to successfully attain their goals. One risk factor is contrasting expectations of the role of the supervisory team. We report a case where we became aware firstly, that a student may have different expectations to us (as supervisors) relating to the level of independence that we expected from them and secondly, that we had unwittingly enabled a passive approach which masked the student’s ability. We subsequently describe a strategy we developed, based on the use of ‘safe words,’ for ensuring that the student’s contributions took centre stage during supervisory meetings. Also considered is how this practice allowed us to make a more accurate assessment of their abilities as well as enabling us to form recommendations to help the student develop their ideas more independently.

Keywords : Autonomy, Supervisory relationships, learner independence

Doctoral research programmes have two functions: a) the generation of new knowledge and b) the development of researchers with accompanying professional identities (Green, 2012). From the theoretical perspective of communities of practice, it is by engagement in the practices of a community that one develops the associated identity of a practitioner (Lave & Wenger, 1991). Yet participation in a research degree programme does not always result in success, either in terms of the generation of new knowledge, or the development of a professional researcher. It is the latter issue that we particularly focus on in the present paper.

Doctoral programme completion rates have been an area of concern for some time (Lindsay, 2015). The quality of supervisory relationships has important implications for doctoral candidates’ attainment of a PhD (Aspland, Edwards, O’Leary, & Ryan, 1999; Lee, 2008; Pyhältö, Vekkaila, & Keskinen, 2015). Consequently, there is a need to enhance the quality of research supervision, particularly as this has implications for research excellence as well as completion rates (Severinsson, 2015). There is a growing awareness of the pedagogic construction of research supervision (Bruce & Stoodley, 2013). Furthermore, the diverse nature of learners joining postgraduate programmes in terms of academic and cultural background and the varied outcomes – few successful candidates will continue to work within academia after graduation – has driven a reconsideration of the requirements of supervision (Harrison & Grant, 2015) alongside government intervention in research higher degrees across the globe (Manathunga, 2005). It is therefore important to consider factors that may impact on the quality of supervisory relationships and how these may be addressed.

Supervising research students can pose a number of challenges. The majority of research examining outcomes for doctoral students has focused upon the role of the supervisory relationship (Bruce & Stoodley, 2013; Lee, 2008; Young, 2014). Research has also identified a range of other issues. Doctoral students are at risk of loneliness and isolation, which has consequences for their psychological wellbeing (Janta, Lugosi, & Brown, 2014). They may be dissatisfied with their supervisors’ knowledge (or lack thereof) about procedures associated with doctoral programmes (Aspland et al, 1999). Doctoral research students also sometimes complain about feedback (e.g. timeliness and quality), yet may fail to accept and respond to feedback (Can & Walker, 2014). The requirements of research students are likely to depend on the stage of their research programme. For example, early stage candidates will tend to need encouragement, guidance and management of expectations (Young, 2014). Students and supervisors may also have different beliefs about the factors that drive doctoral success (Pyhältö et al, 2015). Supervisors are likely to display variations in supervisory style, influenced by factors such as their attachment style (Riggs & Bretz, 2006). Attachment styles can be thought of as psychological constructs representing an individual’s typical pattern of relating to others, e.g. secure relationships, fears that the self is unworthy, or reluctance to form close relationships with others (Bartholomew & Horowitz, 1991).

Contrasting expectations about the nature of supervisory relationships on the part of student and supervisors also contribute to poor student experiences (Harrison & Grant, 2015). For example, students and supervisors may differ in whether they expect a ‘mentorship relationship,’ where the supervisor treats the student as a peer with a shared responsibility for their development, or an ‘apprenticeship relationship’ characterised by the student adopting a passive role to be directed by the supervisor (Wang & Li, 2011). Schemas – cognitive structures reflecting core beliefs about the self, others, and the world – are activated in order to help individuals make sense of life events (Hawke & Provencher, 2011). Where supervisors and students have different schemas relating to supervision, it is likely that they will reach different understandings about events during doctoral programmes, such as the outcome of discussions in supervisory meetings, and may consequently form different interpretations of each other’s behaviours and cognitions. In short, we cannot take for granted that students and supervisors expect similar things from the supervisory relationship and insight is required in order to identify and resolve any misunderstandings. In the present paper, we report a case study of supervision with a student whose particular challenges inspired us to adopt a new pedagogic strategy to accomplish this.

Murphy, Bain and Conrad (2007) identified two broad distinctions in orientations to the supervision of research students: a) the orientation of supervisors to direct the research versus guiding the research process, and b) the orientation to focus supervision on tasks to be accomplished versus the personal development of the research student. In this taxonomy, we position ourselves as supervisors who aim to guide rather than direct in order to foster autonomy, and who prioritise the development of the student. Nonetheless, it should be acknowledged that there are still elements of task focus in our supervision, as the tasks faced by the research student are key to their development as independent researcher and are time constrained by externally‐specified milestones and deadlines. There is therefore some pressure on the supervisory team to keep the student on track, while facilitating autonomy and ‘ownership’ of the research.

Initially we did not have any concerns about the research student who is the subject of this case study. They had previously completed a masters’ level research programme and appeared to have some awareness of research techniques. Later on, we both became aware that the student did not possess the skills or understandings that are required to achieve success at doctoral level. Manathunga (2005) identified four warning signs that are indicative of research students experiencing difficulties in progressing their research: 1) Frequent changing of topic or planned work; 2) Avoiding communications with supervisors; 3) Becoming isolated (from the research environment and peers); and 4) Avoiding submitting work for review. Interestingly, none of these warning signs were present in our student: they regularly communicated with us, and sought frequent review of their work. In fact, we became aware that the student expected a style of feedback that we were not prepared to deliver. An early concern was that the student seemed to desire a passive role and rejected our attempts to encourage a more active orientation to learning. They expected to be directed to the key literature and even to be told which (competing) definition of theoretical constructs they should use. Drawing on Wang and Li’s (2011) work, we sought to develop a mentorship model but soon became aware that we had different expectations of supervision from the student, which we attempted to address during meetings using a range of techniques. Although of some concern, we would typically view this situation as a challenge to work through in order to aid the student’s development of autonomy rather than an immutable failing.

As the first few months of the PhD progressed, we started encouraging our student to work towards the development of a concrete plan for their research programme. It was at this stage that we recognised that the student had insufficient knowledge of key theoretical constructs. The student produced models to be tested that were not underpinned by theory or prior evidence, or a clear rationale for the models and approaches suggested. These issues combined with a lack of detail resulted in vague research plans that were poorly aligned with the stated research goals. This took us by some surprise, as our initial impressions had been that the student had sufficient initial knowledge and abilities to allow them to progress. Contributing to our initial positive impression was their status as a mature student with experience of the professional field to be investigated. This may have encouraged a perception of professional authority in the student. This awareness of the disparity between expectation and subsequent reality led us to reflect on the processes taking place within this supervisory relationship. As active researchers in the students’ field of study who work together closely, we found it all too easy to get excited by discussing the possibilities inherent in the research topic which would extend knowledge

in various ways. During supervisory meetings, we would often identify potential variables or theories that could be relevant, as well as suggest techniques for generating knowledge in the field of study. While we find this kind of exchange unproblematic in our typical working context – groups of researchers and students with sufficient expertise to speak as equals – we came to realise that this approach, rather than inspiring action, masked the challenges experienced by our student. We were suggesting a number of alternative approaches to addressing the research problem, but later realised that our student took all our ideas on board while being seemingly unable to synthesise them to build a rationale for the programme of research, identify a new approach to the problem area, or choose relevant variables to operationalise in their research. This became apparent as the student requested considerable support in making basic research decisions, which failed to meet our expectations for the level of study. This realisation led to a more focused probing during research meetings to identify the extent of the student’s understanding. As described previously, we soon became aware that the student’s awareness of key issues within their chosen field, and the strategic element of relevant research methodology was minimal, and that this has been masked by our previous enthusiasm and readiness to make suggestions.

We therefore decided that in order to evaluate the student’s abilities, but also foster their autonomy, we would need to adopt a new strategy. Key to this strategy was creating ‘space’ within supervisory meetings where the student would be free to speak without our input. We, as supervisors, recognised that we were effectively complicit in the present situation and therefore needed to modify our own behaviours to move forward. We selected a ‘safe word’ methodology in order to warn each other if we became too directive in discussions with the student. This was a strategy adopted from a non‐academic source, specifically the practice of BDSM (Bondage, Domination, Sadism and Masochism). Given that BDSM involves explicit social power dynamics and complex understandings of consent (Barker, 2013), safe words are used as a mechanism to raise awareness of discomfort and immediately stop an activity (Williams, 2006). The use of safe words is also sometimes recommended in couples’ counselling to stop either party saying things during arguments that they may later regret (see Lickerman, 2012). In the meetings with our student, we adopted this technique to avoid ‘helping’ the student (while actually undermining the developmental purpose of the supervisory process).

The benefit of this method was that we could warn each other while not alerting the student, which could potentially have impacted on the student’s confidence. Self‐efficacy, the belief that one can achieve a given task, is crucial in performance (Bandura & Locke, 2003; Bresó, Schaufeli, & Salanova, 2011). When an activity draws attention to a task domain in which an individual perceives low self‐ efficacy, they are less likely to perform well compared to when less attention is drawn to the task domain (Desrichard & Köpetz, 2005). Individuals lower in self‐efficacy are more likely to focus on threatening stimuli rather than positive stimuli (Karademas, Kafetsios, & Sideridis, 2007). In other words, when people lack belief in their ability to achieve, they may focus more on risks than on opportunities. This might apply equally to both supervisors and students! As the student already lacked self‐confidence as an autonomous learner and was more comfortable in a passive role, it was therefore critical that we did not signal our monitoring behaviour, which we believed would disrupt the student’s progress. Further, it aided our goal of observing how the student would perform with minimal influence from us.

We found that, at first, we needed to be conscious in monitoring ourselves and each other, and indeed made use of a safe word. However, over a series of meetings, we found that the practice of developing boundaries around times within meetings when we would offer potential suggestions, after the student had first discussed their perspectives as fully as they were able, became more ingrained. As a consequence, it became easier for us to identify areas in which the student was struggling, that we could then explore more fully. This process allowed us to speak about our concerns openly, which also appeared to be shared by the student. In turn, we were able to discuss and set explicit goals for the student to pursue to demonstrate the level of autonomy expected and the skills and knowledge required for developing a programme of research. We have subsequently reflected on the risks of failing to identify our concerns early on in this research supervision relationship. Specifically, our inability to identify gaps in the student’s knowledge and skills may mean that they approach key transition points (e.g. transfer from MPhil to PhD) being ill equipped to succeed. Failure to identify such gaps means missed opportunities to support students in developing necessary knowledge, skills, and understandings. In the worst cases, this may lead a student to ‘waste’ time and other resources (e.g.finances) in work that is unlikely to produce desirable outcomes. We regard this as a significant ethical issue, as for many students these are not trivial resources.

We recognise in line with comments from a reviewer of an earlier version of this paper that our strategy depends upon a supervisory team. We have focused upon the supervision of research students, in which case there would be a supervisory team. However, there may be times when meetings are held in which only one supervisor may be present. Some readers may also have concerns about issues of student knowledge and independence in cases where they are the sole supervisor (e.g. for dissertations on taught courses). In these cases, the use of a reflective journal may help to surface concerns with student performance. We suggest some reflective questions that may help the independent supervisor:

  • How many ideas (and what ideas) did the student input? Are these ideas clearly focused and articulated?
  • What kind of questions did I ask the student?
  • How many recommendations did I make? How substantive were these recommendations?
  • Did the student express an opinion that I disagreed with? What happened next? How did both student and supervisor deal with the situation?
  • Did I give the student opportunities to challenge me? What did I do to show that it is acceptable to disagree with me?
  • Did I give the student opportunities to reflect back relating to agreements made in the meeting, to ensure mutual understanding?
  • If the student was quiet, or struggled to express an idea, what did I do? Am I comfortable with being quiet while a student seems to struggle? What strategies can I use to manage this better?

In conclusion, we found that by speaking too much during supervisory meetings, we were doing a disservice to our student by reinforcing their preference to take a passive role. By choosing a remedial strategy which would enforce our own silence and effectively ‘police’ each other’s contribution and, at times, dampen down our enthusiasm, we provided our student with the space to demonstrate knowledge, skills and approaches. This allowed us to identify areas of concern more effectively, and to develop measures to help the student to attain key skills and knowledge. Therefore, we recommend the ‘safe word’ strategy for other supervisory teams who are conscious that they may be contributing too much in meetings with research students. We suggest that this may be a particularly useful tool to adopt where there is a risk of assuming that a student has the requisite knowledge and skill, e.g. due to professional background or other forms of personal experience. For example, an individual may have relevant experience, but yet fail to reflect effectively upon the meaning of that experience. There may be other situations in which the use of safe words may be effective, for example to identify a topic that should be avoided, e.g. if one member of the supervisory team focuses too much on late‐stage processes in meetings with a new doctoral student, or topics peripheral to the students’ research which may lead to conflict between supervisors. We do not, however, recommend that this tool be used to mask differences of opinion in supervisors about the research process, as it is important that the research student take a lead in transparent discussions about this. Finally, we note that while this may be an effective tool for managing particular issues in supervisory relationships, the use of safe words must still be contingent upon a good working relationship between supervisors – otherwise there can be no agreement on their use. We also note that it is important to agree the terms of the supervisory approach early with students. In short, safe words provide a tool for managing relationships, but they are not a substitute for forging effective relationships. Similarly, in our own experience, safe words facilitated the surfacing of concerns, but did not replace the need to develop a plan for moving forward.


Aspland, T., Edwards, H., O’Leary, J., & Ryan, Y. (1999) Tracking new directions in the evaluation of postgraduate supervision, Innovative Higher Education, 24 (2), 127‐147

Bandura, A. & Locke, E.A. (2003) Negative self‐efficacy and goal effects revisited, Journal of Applied Psychology, 88 (1), 87‐99

Barker, M. (2013) Consent is a grey area? A comparison of understandings of consent in Fifty Shades of Grey and on the BDSM blogosphere, Sexualities, 16 (8), 896‐914

Bartholomew, K. & Horowitz, L.M. (1991) Attachment styles among young adults: A test of a four‐ category model, Journal of Personality and Social Psychology, 61 (2), 226‐244

Bresó, E., Schaufeli, W.B., & Salanova, M. (2011) Can a self‐efficacy‐based intervention decrease burnout, increase engagement, and enhance performance? A quasi‐experimental study, Higher Education, 61 (4), 339‐355

Bruce, C. & Stoodley, I. (2013) Experiencing higher degree research supervision as teaching, Studies in Higher Education, 38 (2), 226‐241

Can, G. & Walker, A. (2014) Social science doctoral students’ needs and preferences for written feedback, Higher Education, 68, 303‐318

Desrichard, O., & Köpetz, C. (2005) A threat in the elder: The impact of task‐instructions, self‐efficacy and performance expectations on memory performance in the elderly, European Journal of Social Psychology, 35, 537‐552

Green, B. (2012) Addressing the curriculum problem in doctoral education, Australian Universities’ Review, 54 (1), 10‐18

Harrison, S. & Grant, C. (2015) Exploring of new models of research pedagogy: Time to let go of master‐ apprentice style supervision? Teaching in Higher Education, 20 (5), 556‐566

Hawke, L. D., & Provencher, M. D. (2011) Schema Theory and Schema Therapy in Mood and Anxiety Disorders: A Review, Journal Of Cognitive Psychotherapy, 25(4), 257‐276

Janta, H., Lugosi, P., & Brown, L. (2014) Coping with loneliness: A netnographic study of doctoral students, Journal of Further and Higher Education, 38 (4), 553‐571

Karademas, E.C., Kafetsios, K., & Sideridis, G.D. (2007) Optimism, self‐efficacy and information processing of threat‐ and well‐being‐related stimuli, Stress and Health, 23, 285‐294

Lave, J. & Wenger, E. (1991) Situated learning: Legitimate peripheral participation, Cambridge: Cambridge University Press

Lee, A. (2008) How are doctoral students supervised? Concepts of doctoral research supervision, Studies in Higher Education, 33 (3), 267‐281

Lickerman, A. (2012) The magical power of ‘safe’ words to prevent harm, Psychology Today, available at https:// ‐in‐world/201211/the‐magical‐power‐safe‐ words‐prevent‐harm

Lindsay, S. (2015) What works for doctoral students in completing their thesis? Teaching In Higher Education, 20 (2), 183‐196

Manathunga, C. (2005) Early warning signs in postgraduate research education: A different approach to ensuring timely completions, Teaching in Higher Education, 10 (2), 219‐233

Murphy, N., Bain, J.D., & Conrad, L. (2007) Orientations to research higher degree supervision, Higher Education, 53, 209‐234

Pyhältö, K., Vekkaila, J., & Keskinen, J. (2015) Fit matters in the supervisory relationship: doctoral students and supervisors perceptions about the supervisory activities. Innovations in Education and Teaching International, 52 (1), 4‐16

Riggs, A.A. & Bretz, K.M. (2006) Attachment processes in the supervisory relationship: An exploratory investigation, Professional Psychology: Research and Practice, 37 (5), 558‐566

Severinsson, E. (2015) Rights and responsibilities in research supervision, Nursing and Health Sciences, 17, 195‐200

Wang, T. & Li, L.Y. (2011) ‘Tell me what to do’ vs. ‘guide me through it’: Feedback experiences of international doctoral students,Active Learning in Higher Education, 12 (2), 101‐112

Williams, D.J. (2006) Different (painful!) strokes for different folks: A general overview of sexual sadomasochism (SM) and its diversity, Sexual Addiction & Compulsivity, 13, 333‐346

Young, K.J. (2014) Research mentoring: Suggestions and encouragement from a reflection exercise, Journal of Chiropractic Education , 28 (2), 168‐172


  • There are currently no refbacks.

Copyright (c) 2017 Andrew James Clements, Gail Kinman