Introduction

The behavior and practices of research supervisors greatly affects the satisfaction and success of graduate researchers, such as PhD candidates.  Indeed, when asked to identify their key problems or obstacles, research candidates often refer to inadequate or fractious relationships with supervisors.  Yet, supervision is a challenging and multi-faceted endeavor.  Supervisors need to fulfill many responsibilities and apply many skills.    

After a comprehensive review of the literature, however, Fulgence (2019) concluded that many of the capabilities that supervisors need to develop can be divided into six main constellations:

• expertise in the academic discipline, such as knowledge about the main theories and approaches, the culture and discourse of this discipline, and the settings or context

• expertise in research skills, such as how to construct a research question, choose a research design, collect, analyze, and interpret data, identify limitations, and communicate the findings in oral presentations and written documents

• the capacity to navigate workplaces effectively, such as skills in teamwork, resilience, problem solving, decision making, critical thinking, adaptability, and leadership

• the ability to manage projects and resources, such as how to manage time, manage finances, and accommodate diversity

• familiarity with digital skills, such as online tools and online networks

• knowledge about the doctoral journey, such as university procedures, supervision roles, feedback and delivery.

Supervisors can distill these skills from a range of experiences and sources.  In particular, supervisors may develop these skills if they

• act as a supervisor, perhaps under the guidance of a principal supervisor (e.g., Halse, 2011)

• receive mentoring and career coaching (Rogers et al., 2018)

• seek information themselves from books, publications, or other supervisors (Gordon, 2014)

• attend supervision training courses or workshops (e.g., Botha & Muller, 2016)

Nevertheless, to help supervisors acquire these skills, institutions are becoming increasingly reliant on training courses or workshops.  Many universities oblige research supervisors to complete formal training and development, designed to enhance their supervisory practices (McCulloch & Loeser, 2016).  National good practice guidelines, such as the Quality Assurance Agency for Higher Education in the UK and the Australian Council of Graduate Research, often mandate this formal training and development of supervisors.  Similarly, the Salzburg Principles Section 2.3, stipulated by the European Universities Association (EUA Council for Doctoral Education, 2010), maintains that institutions are assigned the responsibility to facilitate the professional development of supervisors.

This training and development of supervisors can range from short workshops, lasting an hour or two, to comprehensive programs.  An example of a comprehensive program is the postgraduate certificate in research degree supervision, introduced by the Edge Hill College of Higher Education in 1999 (Cryer, 2000).

As Wisker and Kiley (2014) underscored, these programs tend to vary along two dimensions.  First, these experiences vary on the duration over which development evolves.  At one pole, most of the development opportunities may be confined to one time, often induction.  At other institutions, the development opportunities might continue indefinitely, sometimes called CPD or continued professional development.

Second, the breadth of content also varies across these development experiences.  Some training and development are confined to topics that are directly relevant to immediate supervision needs, such as the HDR policies and procedures and some guidelines that supervisors can apply now.  Other training and development programs are more encompassing and impart knowledge and skills that could be relevant to the future, such as the changing landscape of doctoral education, the pedagogy of graduate research, and reflections on case studies.   

Besides these two dimensions, training programs also vary on whether they are mandatory.  Many institutions grapple with the tension between mandatory programs in which participants often feel resentment or show resistance and optional programs in which participants often do not attend (Manathunga, 2005).  Supervisors who are not reflective, and thus often ineffective, may be the least inclined to attend.

As McCulloch and Loeser (2016) bemoan, few studies have substantiated the benefits of these initiatives.  Although institutions often evaluate this training and development of supervisors—perhaps by asking participants to complete evaluation forms, for example—this evaluation does not necessarily indicate whether the event actually improved supervisory behavior and benefited graduate researchers. 

Examples of supervision training: A program at Coventry University

Although not prevalent, scholars occasionally publish journal articles that delineate and evaluate a supervision training program.  To illustrate, Jara (2021) designed and commenced an evaluation of a program that facilitates the development of research supervisors.  This study, in essence, revealed that supervisors greatly appreciated the discussion of case studies, the exchange of experiences, the insights of experienced supervisors, and the reflection on their own practice.  Specifically, a team of staff at Coventry University reviewed the literature to determine the contents of this program, comprising a set of nine independent modules they could complete in any order, such as

• the role of supervisors

• how to support diverse candidates

• how to deliver effective feedback

• how to improve the writing and critical thinking of candidates

• how to monitor progress

• how to navigate the examination process

• how to maintain effective and trusting relationships

During each module, participants read a relevant empirical article in advance and discussed key issues that were relevant to this paper, lasting about four hours. These discussions revolved around hypothetical case studies, exchange of experiences, insights from experienced supervisors, and reflection on practice. The cases were designed to reveal multifaceted complications in diverse settings.  These case studies, typically fewer than 300 words, were derived from personal experiences, books, articles, and websites, such as academia.stackexchange.com.  The modules were conducted online, but individuals could attend in person.

To evaluate the program, participants completed a survey at the end of each module and could participate in interviews after the program ended.  Overall, the results were encouraging:

• about 63% of participants completed all the modules, a reasonable level of attendance, suggesting this modular approach may improve participation. A previous variant of this program that was not modular did not attract this level of participation

• the vast majority of participants—over 90%--agreed or strongly agreed with statements that indicate the program was coherent, the discussions and activities facilitated learning and reflection, the insights could be applied in practice, and the experienced improved their confidence

• participants especially appreciated the case studies that depict the challenges of PhDs from the perspective of candidates

• however, only 70% of the participants agreed or strongly agreed the concept or strategies were new. 

These findings indicate that participants can perceive these experiences as invaluable and helpful to their supervisory practice, but even if they do not learn novel concepts or strategies. 

Examples of supervision training: The supervising at UniSA workshop

Rather than comprehensive programs, some universities present shorter courses, such as induction training, that is designed to impart key insights but also inspire supervisors to develop their practices over time in the future.  For example, McCulloch and Loeser (2016) outlined and evaluated the supervision program at the University of South Australia.  This induction workshop introduces supervisors to doctoral education at the university and imparts some insights on the pedagogy of supervision as well as inspires supervisors to pursue more training and development in the future.  The workshop is conceptualized as an opportunity to reflect upon supervision and collaborate with peers rather than as a hurdle that teach these individuals how to supervise.  Hence, the workshop primarily adopts a constructivist approach to learning, in which supervisors develop their own insights, rather than merely a didactic approach—although some information is disseminated.  The workshop, offered several times a year, lasts one day, including lunch, during which participants eat lunch together and engage with each other. 

Before lunch, the workshop comprises two sections.  First, the Dean of Graduate Studies and administrative staff from the research team briefly outline the policies, procedures, and operations at the university that relate to graduate research as well as the key trends across the nation.  Second, a specialist in the field demonstrates how doctoral education has shifted over the years and highlights the key challenges that supervisors are increasingly experiencing.

After lunch, the workshop also comprises two sections.  Specifically, experienced research supervisors at the university first discuss their solutions and responses to many of the challenges that supervisors are increasingly experiencing.   And finally, a specialist in the field invites small teams of supervisors receive case studies, delineating common problems that candidates and supervisors experience.  These teams develop strategies to resolve these problems—and then share these strategies with the other participants in the workshop. 

After participants complete this workshop, they receive a digital badge, embedded in a platform that recognizes all the development activities that research candidates and supervisors complete.  They also complete an evaluation sheet immediately after the workshop as well as a more extensive questionnaire months or years later, designed to assess the lasting impact of this experience.  For example, in addition to demographic questions, the questions prompt the supervisors to consider which insights and practices they had applied or suggested to peers and whether they had pursued other training and development activities. 

In general, the results of this questionnaire indicated the workshops were useful.  Specifically

• more than 50% of participants felt the workshop enhanced their understanding of the national trends around graduate research, how graduate research has shifted in the last two decades, the university policies and procedures, and the online support that is available

• for example, 81% of participants felt the workshop enhanced their understanding of the national trends either to a large extent or to some extent

• likewise, more than 50% of participants felt they had applied the knowledge and insights they learned, especially around university policies, university procedures, scholarships, online resources, and the strategies that experienced supervisors had applied

Examples of supervision training: A program directed at senior or experienced supervisors

Many supervisor training programs are primarily designed to benefit inexperienced supervisors (for an exception, revolving around conversations, see Spiller et al., 2013).  In contrast, Wichmann-Hansen et al (2020) delineated a program, launched in a Danish university, lasting an extended duration, but targeted more at experienced supervisors. Thirteen senior academics participated.  

Participation was mandatory and embedded within the job description.  If individuals do not complete the course, some funding is withheld from the budget; otherwise, some funding is directed to the budget.  To receive this incentive, participants needed to complete all homework activities and attend the sessions, consuming about 60 hours altogether.

To evaluate this program, Wichmann-Hansen et al (2020) derived data from evaluation forms that participants completed after the program ended, comprising rating scales and short answers, as well as perceptions of candidates on whether the quality of supervision had improved.   

The program mainly revolved around how to deliver constructive feedback to drafts, how to maintain trusting relationships with candidates while they still monitor progress, and how to promote the independence of candidates.  During the sessions,

• the participants interacted in groups to discuss issues and cases

• the participants completed quizzes, with MentiMeter to monitor feedback

• the participants listened to insights about best practice

​

About half the 60 hours was dedicated to homework activities.  The homework activities included

• uploading their feedback on student drafts

• writing their MoU or supervision agreement, on the roles and responsibilities of supervisors and candidates, and delivering feedback to other MoUs

• write about a challenging case in supervision and then to share this case later in the workshop

• writing a scholarly essay on the lessons participants gained from their observations of other supervisors; these observations were video recordings of actual meetings

The analysis suggested the program was generally effective.  Specifically

• 93% of participants rated the program as effective or very effective

• participants valued the opportunity to share practices and tools

• participants, although skeptical initially, greatly appreciated the opportunity to receive feedback about their recorded practices

• participants felt that, in the future, they will ask more open-ended questions to candidates, listen actively, clarify expectations, and implement better techniques to deliver feedback

• more than 80% of candidates were more satisfied with their supervisor after this training

Challenges of candidates that could inform case studies and experiences: Liminal experiences

As these examples and many other studies demonstrate, supervisors often benefit from reflections and discussions about case studies and the challenging that peers experience.  Therefore, to improve supervision training and development, coordinators need to understand some of the challenges that candidates, supervisors, or both candidates and supervisors experience.    

Keefer (2015) invoked the notion of liminality to characterize some of the challenges that candidates experience.  That is, during graduate research, candidates often experience sudden insights or shifts that transform the perceptions or understandings of some topic or matter.  Typically, this transformation follows a period of uncertainty or confusion, called a liminal state. Liminality refers to the state after individuals have detached themselves from a previous identity or perspective but have yet to adopt the transformed identity or perspective entirely.  These states are common in doctoral candidates, who often waver between their identity as students and their identity as academics, often manifesting as imposter syndrome.  During most of their journey, these candidates experience obstacles and troubles that challenge their identity as academics.  The role of supervisors is partly to help candidates circumvent these obstacles and assume the role of academics. 

Keefer (2015) conducted a narrative inquiry to characterize these challenges and liminal experiences in 23 doctoral candidates.  The aim of this study was to help supervisors appreciate the sources of this uncertainty and confusion in doctoral candidates and, thus, address these liminal experiences.    

The results uncovered three key liminal experiences or challenges that many candidates, across a range of disciplines, experience.  These experiences can last days, weeks, or even months and years.  The first experience revolves around feelings of isolation and loneliness.   Candidates referred to feeling of solitude, amplified by the unique path that each individual must travel to unearth original findings.  Participants also referred to the burden they feel knowing they are entirely responsible for their research.  These individuals did not expect to experience this loneliness, partly because they heard about the success of peers—success that only magnified their sense of loneliness.

The second liminal experience revolved around imposter syndrome, in which candidates questioned whether they have acquired the skills and capabilities to thrive in academia.  Some participants were reluctant to disclose these concerns to their supervisor, partly because of unpleasant experiences in the past.  Indeed, candidates often experience an incessant feeling their incompetence would eventually be unmasked. They felt their peers would recognize they do not belong.  Explicit references to imposter syndrome helped some participants, because they felt the term normalizes the experience.   

The final liminal experience revolved a sense of misalignment between the candidate and the supervisor or institution on research priorities and practices.  Sometimes, this misalignment emanated from differences in the background, experience, qualifications, and personality between candidates and supervisors or other academics.  But this misalignment can manifest in candidates as serious doubts about their capabilities and direction, after they receive strident, albeit, unfair criticism from supervisors or other academics.  

The implication of these findings, according to Keefer (2015), is that supervisors should forewarn candidates about these potential challenges and offer to discuss these matters if they transpire.  Supervisors might also suggest other avenues candidates may pursue in response to these concerns. 

Challenges of supervisors that could inform case studies and experiences: Admission

Supervisors experience a range of challenges across the entire doctoral journey, from admission to examination.  To illustrate challenges during admission, in study of 20 supervisors and 20 doctoral candidates, conducted by Denis et al. (2019), supervisors conceded they could not really predict which applicants are likely to thrive during their degree. Despite this uncertainty, inexperienced supervisors did not feel they were granted the luxury to reject many applicants—because they cannot as readily attract this interest.

Challenges of supervisors that could inform case studies and experiences: Remote supervision

When supervisors and candidates are not located in same region, and thus must communicate remotely, other challenges are unlikely.  Nasiri and Mafakheri (2015) conducted a comprehensive review to identify, and to characterize, these challenges as well as to unearth the strategies that supervisors apply to address these challenges.   As examples of challenges, the authors revealed that

• if located in separate time zones, supervisors and candidates need to identify a time that suits both parties.  This goal may be hard, because individuals often thrive on specific activities, such as activities that demand creativity, at particular times of the day (Breslin, 2019).

• If supervisors and candidates do not meet in person, they may not feel a sense of familiarity with one another; the conversations, therefore, tend to be more formal and stilted rather than engaging and motivating (cf Sussex, 2008).   

• If located in separate time zones, supervisors might receive messages from a candidate outside their usual work hours.  Responses to many emails outside hours—or even the prospect

Challenges of supervisors that could inform case studies and experiences: Changes over time

The practice of supervision has gradually, and sometimes abruptly, shifted in the last few decades.  For example

• in some previous decades, candidates were usually assigned only one supervisor; today, the vast majority of candidates are assigned two or more supervisors, usually operating as a panel (Green & Bowden, 2012)

• in previous decades, supervisors often perceived themselves as the master, whose role was to impart knowledge to their apprentice; today, many candidates now prefer and expect their supervisor to facilitate their development collaboratively rather than merely impart knowledge or convey instructions (Fenge, 2012).  

Other important content of supervision training: Technology

Although supervisors benefit if they analyze cases, share experiences, and reflect upon their practices, institutions still need to confirm that vital content is included in these workshop. Maor et al. (2016) argued that perhaps more training and opportunities around technology, such as Web 2.0 tools, could also facilitate supervision. 

This recommendation was derived from the research that Maor et al conducted about Web 2.0 and online communities in doctoral research.  Specifically, Maor et al conducted a comprehensive review to explore this topic.  Initially, 196 papers that had been peer reviewed, 64 conference proceedings, 8 theses, and 16 reports fulfilled the selection criteria.  The authors then extracted the 18 most relevant empirical articles on this topic and derived the recommendations from these sources.  These articles examined or alluded to range of possible on technologies.  These technologies include

• tools that facilitate communication over distance, such as Elluminate, Wimba, and Second Life

• tools that enable individuals to share information and insights, such as Microblogging and e-Portfolios

• more common tools such as Skype and Microsoft Office SharePoint.

Furthermore, in two of the studies, the institutions had developed their own virtual spaces: Doctoralnet (Danby & Lee, 2012) and Form@doct (Malingre et al., 2013).

The main insight from these studies was the advent of a novel pedagogy, revolving around more sustained virtual communities of candidates and sometimes their supervisors, interested in an overlapping topic, in which individuals learned from one another and debated a range of relevant issues.  Individuals might initially observe these communities.  But then gradually, as they develop specialist knowledge and use tools that facilitate knowledge exchange—such as ePortfolios—they become more active participants.  When these communities were most effective, rather than adopt the role as elite specialists or masters, the supervisors were often collaborative participants of these ecosystems.  Doctoral candidates were granted more autonomy on when and how to learn.  

Several tools, applications, platforms, and websites, such as Doctoralnet, can underpin these communities.  These tools enable candidates and supervisors to discuss issues, across institutions, collaborate on writing, and share information. 

Yet, in practice, many supervisors were not as familiar as their candidates in Web 2.0 technologies and, therefore, continued to adopt more traditional approaches.  These supervisors did not often encourage their candidates to engage with online communities.   

Even if supervisors contribute to these online communities, these collaborations may subside over time. To sustain these communities, coordinators may need to attract significant individuals, such as renowned academics or potential employers, to contribute occasionally. 

Other sources of development

Formal training sessions are not the only source of development in supervisors.  Indeed, Fulgence (2019) classified all the experiences and sources of this development, derived from past studies, into five clusters.  First, supervisors might engage in deliberate education and training to develop these skills.   They might attend workshops, either at their institution or outside their institution, both online and in person, or read articles and books on supervision.  Some MOOCs, including Futurelearn and AuthorAID, impart supervision skills.  Alternatively, supervisors might have even learned some of these skills during the training they received as a graduate candidate (Reguero et al., 2017).    

Second, supervisors might derive these skills from their experiences in the realm of research.  That is, during their doctorate or other research experiences, individuals are exposed to the practices and beliefs of many individuals, such as their supervisors or contacts the meet at conferences.  These practices and beliefs shape their own knowledge and assumption about suitable research practices in general and supervision practices in particular.  That is, during these experiences, they learn a range of management skills, research knowledge, supervision practices, networking abilities, and interdisciplinary skills, all of which shape their supervision practices.   Indeed, according to Henderson, (2018), the practices of inexperienced supervisors are largely dependent on their practices of their supervisor during the doctoral studies.

Third, when academics are finally granted opportunities to supervise, this experience alone can shape their supervisory capabilities and tendencies (Reguero et al., 2017).  They might derive insights from their observations of other supervisors on the panel, from reflections on their behavior, or from the advice of mentors they might arrange to facilitate this supervision.  These experiences can help supervisors acquire more extensive knowledge about their discipline, research practices, relationship development, time management, and pedagogical practices.  

Yet, these experiences, unless buttressed with more formal opportunities, is not sufficient to facilitate the development of supervisors.  The fourth source of supervisor development—the policies, procedures, and guidelines of institutions—can overcome this shortfall, at least partly.  These policies not only impart key information, such as the role and responsibilities of supervisors, but embed practices that can also be a source of learning, such as the feedback of examiners. 

Finally, supervisors might participate in research that is designed to improve supervision.  They might collate, distil, and analyze data that could be relevant to their role—such as the concerns that other candidates have raised about their supervision.   

References

• Bitzer, E. M., Trafford, V. N., & Leshem, S. (2013). 'Love it when you speak foreign': A trans-national perspective on the professional development of doctoral supervisors in South Africa. South African Journal of Higher Education, 27(4), 781-796.

• Bøgelund, P. (2015). How supervisors perceive PhD supervision – And how they practice it. International Journal of Doctoral Studies, 10, 39-55.

• Botha, J., & Muller, N. (Eds.) (2016). Institutional research in South African higher education. Stellenbosch: SUN PRESS

• Breslin, D. (2019). Group creativity and the time of the day. Studies in Higher Education, 44(7), 1103-1118.

• Danby, S., & Lee, A. (2012). Researching doctoral pedagogy close up: Design and action in two doctoral programmes. Australian Universities' Review, The, 54(1), 19-28.

• Denis, C., Colet, N. R., & Lison, C. (2019). Doctoral supervision in North America: Perception and challenges of supervisor and supervisee. Higher Education Studies, 9(1), 30-39.

• Dimitrova, R. (2016). Ingredients of good PhD supervision–Evidence from a student survey at Stockholm University. Utbildning & Lärande, 10(14), 40-52.

• Fenge, L. A. (2012). Enhancing the doctoral journey: The role of group supervision in supporting collaborative learning and creativity. Studies in Higher education, 37(4), 401-414.

• Fulgence, K. (2019). A theoretical perspective on how doctoral supervisors develop supervision skills. International Journal of Doctoral Studies, 14, 721.

• Gordon, N. (2014). Flexible pedagogies: Technology-enhanced learning. Heslington, UK: The Higher Education Academy.        

• Green, P., & Bowden, J. (2012). Completion mindsets and contexts in doctoral supervision. Quality Assurance in Education, 20(1), 66–80.

• Guerin, C., Kerr, H., & Green, I. (2015). Supervision pedagogies: Narratives from the field. Teaching in Higher Education, 20(1), 107-118.

• Halse, C. (2011). Becoming a supervisor: The impact of doctoral supervision on supervisors’ learning. Studies in Higher Education, 36(5), 557-570.

• Halse, C., & Bansel, P. (2012). The learning alliance: Ethics in doctoral supervision. Oxford Review of Education, 38(4), 377-392.

• Henderson, E. F. (2018). Anticipating doctoral supervision: (Not) bridging the transition from supervisee to supervisor. Teaching in Higher Education, 23(4), 403–418.

• Hill, G., & Vaughan, S. (2018). Conversations about research supervision – Enabling and accrediting a community of practice model for research degree supervisor development. Innovations in Education and Teaching International, 55(2), 153–163.

• Jara, M. (2021). Research-based doctoral supervision development programme: Learning through peer learning, reflection and case studies. Innovations in Education and Teaching International, 58(4), 441–450.

• Keefer, J. M. (2015). Experiencing doctoral liminality as a conceptual threshold and how supervisors can use it. Innovations in Education and Teaching International, 52(1), 17-28.

• Lee, A. (2018). How can we develop supervisors for the modern doctorate? Studies in Higher Education, 43(5), 878-890.

• Malingre, M. L., Serres, A., Sainsot, A., & Men, H. L. (2013). Form@ doct: Designing innovative online tutorials for PhD students in France. IFLA journal, 39(1), 45-57.

• Manathunga, C. (2005). The development of research supervision: “Turning the light on a private space”. International Journal for Academic Development, 10(1), 17-30.

• Maor, D., Ensor, J. D., & Fraser, B. J. (2016). Doctoral supervision in virtual spaces: A review of research of web-based tools to develop collaborative supervision. Higher Education Research & Development, 35(1), 172-188.

• McCallin, A., & Nayar, S. (2011). Postgraduate research supervision: A critical review of current practice. Teaching in Higher Education, 17(1), 63-74.

• McCulloch, A., & Loeser, C. (2016). Does research degree supervisor training work? The impact of a professional development induction workshop on supervision practice. Higher Education Research and Development, 35(5), 968–982.

• Nasiri, F., & Mafakheri, F. (2015). Postgraduate research supervision at a distance: A review of challenges and strategies. Studies in Higher Education, 40(10), 1962-1969.

• Olmos-López, P., & Sunderland, J. (2017). Doctoral supervisors’ and supervisees’ responses to co-supervision. Journal of Further and Higher Education, 41(6), 727-740.

• Pearson, M., & Kayrooz, C. (2004). Enabling critical reflection on research supervisory practice. International Journal for Academic Development, 9(1), 99-116.

• Reguero, M., Carvajal, J. J., García, M. E., & Valverde, M. (Eds.) (2017). Good practices in doctoral supervision: Reflections from the Tarragona Think Tank. Tarragona, Spain: Universitat Rovirai Virgili.

• Rogers, J., Sorkness, C., Spencer, K., & Pfund, C. (2018). Increasing research mentor training among biomedical researchers at Clinical and Translational Science Award hubs: The impact of the facilitator training initiative. Journal of Clinical and Translational Science, 2(3), 118-123.

• Saleem, T., & Mehmood, N. (2018). Assessing the quality of supervision experiences in the different research stages at postgraduate level. Journal of Education and Educational Development, 5(2), 8-27.

• Spiller, D., Byrnes, G., & Bruce Ferguson, P. (2013). Enhancing postgraduate supervision through a process of conversational inquiry. Higher Education Research & Development, 32(5), 833-845.

• Sussex, R. (2008). Technological options in supervising remote research students. Higher Education, 55(1), 121-137.

• Sweitzer, V. B. (2009). Towards a theory of doctoral student professional identity development: A developmental networks approach. The Journal of Higher Education, 80(1), 1-33.

• Turner, G. (2015). Learning to supervise: Four journeys. Innovations in Education and Teaching International, 52(1), 86-98.

• Vilkinas, T. (2008). An exploratory study of the supervision of PhD/research students’ theses. Innovative Higher Education, 32(5), 297-311.

• Wichmann-Hansen, G., Godskesen, M., & Kiley, M. (2020). Successful development programs for experienced doctoral supervisors – What does it take? International Journal for Academic Development, 25(2), 176-188.

• Wisker, G., & Kiley, M. (2014). Professional learning: lessons for supervision from doctoral examining. International Journal for Academic Development, 19(2), 125-138.