PhD students, significant others, and pedagogical conversations. The importance of trusting relationships for academic development

ABSTRACT

This study explores how trusting relationships between PhD students and their significant others contribute to the success of academic development programs, and how these programs affect trust in significant others. It introduces a definition of trust and distrust in academic development as synthesized from trust research in other disciplines. The study finds that lack of trust hinders conversations in general and about certain topics in particular. Academic development programs can improve participants’ trustworthiness by positioning them as expert conversation partners; but they may also diminish participants’ trust in significant others if PhD students’ new role conception differs from their significant others’.

KEYWORDS:

• Academic development program
• phd students
• trust
• trusting relationships
• pedagogical conversations
• role theory

Introduction

Previous studies (among them Pataraia et al., 2014; Roxå & Mårtensson, 2009; Thomson & Trigwell, 2018) have examined the nature and purpose of pedagogical conversations for academic development. Focusing primarily on discussion networks (Benbow & Lee, 2019; Pataraia et al., 2014; Rienties & Hosein, 2015; Roxå et al., 2011; Van Waes et al., 2015), these studies have established the importance of conversations between colleagues for learning. Some have advocated for trust and trust-development, arguing that trust is vital for conversations to take place and for innovative teaching methods to spread within an institution (Chng & Geertsema, 2016; Kezar, 2004; Murray, 2016; Stocks & Trevitt, 2016). The empirical evidence presented by Pataraia et al. (2014), Rienties and Hosein (2015), and Roxå and Mårtensson (2009, 2015) lead to similar conclusions. However, we still know very little about how trust and pedagogical discussion relate to each other as no previous study, to our knowledge, has explicitly set out to explore this relationship.

This study fills this gap by asking how trusting relationships between PhD students and their significant conversation partners contribute to the success of academic development programs and how academic development programs affect trust. Our research, which draws on symbolic interactionist role theory (Charon, 2001; Mead, 1934), investigates trust in those relationships from the point of view of PhD students who participated in an academic development program. Using interview data from both program cohorts (2017–2018, 2018–2019), we analyze the teaching-related conversations of 21 PhD students (16 from Masaryk University and 5 from the University of Economics in Bratislava) and the person they named as their most significant other (Charon, 2001, p. 76) regarding their teaching. Traditionally, PhD students have less choice about whom they can talk to, as they must converse with certain faculty members. This makes it even more important to understand their perceptions, feelings, and behavior towards their conversation partners in order to uncover how such relationships impact the effectiveness of academic development programs.

This article contributes to the extant literature in two ways. Firstly, it goes beyond our current knowledge on trust by introducing findings on trust from other disciplines and by exploring the nature of the relationship between (dis)trustful conversations and the success of academic development programs. Secondly, this article advances our understanding of the link between trust and pedagogical discussions by exploring interpersonal trust via individual relationships, rather than continuing to explore relationship networks of participants of academic development programs (for example, Pataraia et al., 2014; Roxå et al., 2011). Focusing on individuals’ relationship approaches, allows for a more in depth look into these relationships. Finally, by introducing non-trust conditions to the academic development literature, it becomes possible to examine trust from a comparative perspective and in contrast to situations in which it is absent.

Teaching-related conversations in trusting and non-trusting relationships

Participating in both academic development programs and in pedagogical conversations with established faculty members are ways in which PhD students advance their pedagogical thinking and practice. In the language of symbolic interactionist role theory, these processes are used to create social reality in terms of interaction between two agents – self and other. These two agents enact roles, that is, social positions, and do so in response to the role-taking of the other as well as the broader situation in which the agents operate. This way, roles comprise behavioral expectations towards both oneself and the other about the forms of behavior they believe to be appropriate in particular situations (Charon, 2001, pp. 114–116; see also Mead, 1934). In this article, we focus on one particular type of role-taking – that of the teacher. Depending on the nature of interactions with significant others, that role may be understood and performed in various ways (Charon, 2001). We investigate the impact of a situational factor – trust or its absence – on significant other interactions to understand the impact of trust on the role definition and role behavior of PhD students.

Individuals understand roles and regulate their own behavior by taking the perspective of certain other individuals, termed ‘significant others’. Significant others are those people ‘who take on importance to the individual, those whom the individual desires to impress; they might be those he or she respects, those he or she wants acceptance from, those he or she fears, or those with whom he or she identifies’ (Charon, 2001, p. 76). This definition shares the interactionist and constructivist underpinnings of Berger’s and Luckman’s (1966) understanding of this concept, which is used in the conversations literature (Benbow & Lee, 2019; Roxå & Mårtensson, 2009, 2015; Roxå et al., 2011). However, Mead/Charon do not require trust in a relationship with a significant other. Understanding trust this way allows for the examination of role-taking and role performance under various trust conditions. It is also congruent with the fact that in a work environment, one may not freely choose who one works with or talks to, which is particularly true for PhD students who stand on the bottom rung of the career ladder.

Such a role theoretical approach sits well with the experience-based and domain-specific nature of trust (Lewicki et al., 1998; Mayer et al., 1995), which we define as a psychological state of an actor (the trustor) who is willing to accept vulnerability to another individual (the trustee) on the basis of positive expectations regarding the intentions and the behavior of the trustee (Mayer et al., 1995). The decision to trust is the result of a combined judgment about another person’s likeability and trustworthiness, where trustworthiness entails an assessment of the other’s ability, integrity, and benevolence.

Moreover, trust is built as one agent, through their interactions, convinces the other of their capabilities, honesty, and benevolence. Conversely, trust dissipates as doubts arise about an agent in one or more of those dimensions. Behaviorally, trust encourages a readiness to take what a third party – but not the trustor – would see as risk-taking (Nikolova et al., 2015). Trust also results in open, frequent, and collaborative patterns of communication in which parties exchange even sensitive information, do not hesitate to ask for help, and coordinate actions (Koeszegi, 2004). When trust is missing from the relationship, an agent expects harm from the other and prefers to avoid interaction altogether (McKnight & Chervany, 2001, pp. 884–885). When interaction is unavoidable, distrustors are vigilant and monitor the other to protect themselves from harmful behavior. Conversations are reduced to the absolute minimum and important information is withheld, or the information conveyed during discussions is often distorted, filtered, or kept to oneself as much as possible (Lewicki et al., 1998; McKnight & Chervany, 2001).

Based on this previous work in the field, we firstly expect that those in a trusting significant other relationship are likely to have more conversations about teaching than those who are not in a trusting relationship. Secondly, we expect that in order to minimize vulnerability in non-trusting working relationships, non-trusting participants are likely to restrict discussions to unavoidable topics such as course content, assessment issues (such as exam scheduling and grading), and administrative issues, while avoiding altogether issues such as teaching methods, syllabus design, the students, and reflection on their own teaching. Thirdly, we expect that participants in a non-trusting relationship will change their existing non-trusting significant other relationship either by choosing a different conversation partner, or by attempting to build more trust into the existing relationship. Finally, we posit that the additional knowledge and skills gained while participating in the academic development program would likely influence the nature of the relationship and, with it, conversations. Our assumption is that PhD students’ increased ability to converse about teaching would be noticed by their significant others, thereby improving the former’s trustworthiness.

Research design

We interviewed participants of an academic development program before they started the program and after program completion. The one-year program, Learning-centered and reflective teaching. From theory to good practice, was a joint program, funded by an Erasmus+ grant, for PhD students from Masaryk University, Brno, Czech Republic, and the University of Economics in Bratislava, Slovakia. The program started with an eight-day, face-to-face training course during which participants were introduced to the principles of good teaching in higher education and the key concepts of university pedagogy. Each participant was then paired with a professional academic developer who served as the participant’s coach in the 9-month online follow-up component of the course. During this later phase, participants were expected to integrate the knowledge and skills developed in the first part of the program with their teaching practice, and to reflect on their teaching experience and the effectiveness of their teaching innovation in the form of a research paper (For more, see Pleschová and Simon, 2021). As outlined in the program’s philosophy of change (Pleschová, 2018), the program was designed to encourage and enable PhD students to discuss teaching and learning with their colleagues, because conversations can play a role in formulating and maintaining participants’ change in thinking and practice in line with the program goals (Roxå & Mårtensson, 2009). Research has also shown that knowledge from academic development programs spreads through participants’ conversations with colleagues not directly involved in such programs (Thomson, 2015).

Because a similar program had reported that reactions from participants’ immediate environment (department heads and other colleagues) hindered them from changing their teaching in line with program goals (Pleschová & McAlpine, 2016), we decided to explore PhD students’ institutional context, their most significant teaching-related relationship, and pedagogical conversations they had with these individuals. Accordingly, interviews started with PhD students being asked to identify the person they found most significant for their teaching practice and development as teachers at the institution where they undertook their PhD studies. Next, we used card sorts to determine the nature of trust in the relationship in a systematic manner, with reasonable reliability (Rugg & McGeorge, 2005), and without sensitizing research subjects to trust. To avoid priming interview subjects to trust, they were only informed of the overall purpose of the research (‘to understand their feelings and behavior with regards to their teaching’), and not its explicit focus on trust. This decision was approved as part of the ethics cover obtained via Central European University.

Overall, we used 57 cards with words describing feelings and behaviors. Roughly half (30) were related to the trust dimension, while the rest (27) described other feelings and behaviors (see Table 1). We asked interview subjects to see the cards as representations of their potential feelings towards their significant other and sort three times to account for the nature and strength of their feelings. After the sorting, the interview continued with an in-depth discussion that, with the help of information gained through the sorting exercise, explored pedagogical conversations and the relationship of such conversations to trust. Post-test interviews contained additional questions, investigating changes in relationships and the causes of these changes (see Rugg & McGeorge, 2005 for details about sorting methods).

Table 1. Feelings and behaviors related to trust conditions and included in the card sorts

Card-sort categories*		Card-sort words/phases
Trust and distrust		trustful, distrustful
Expressions and manifestations of trust		confident, faithful, hopeful, safe, supported, encouraged, comfortable, active, take the initiative, feeling to have a choice, frequent interaction
Expressions and manifestations of lack of trust*	Distrust	afraid, anxious, skeptical, cynical, watchful for harm, withholding information, passive, avoid interaction, powerless
	Ambivalence	hesitant, suspicious, confused, uncertain, surprised, monitoring, insecure, ambivalent
Other		angry, calm, cheerful, concerned, demoralized, depressed, determined, disinterested, eager, enthusiastic, excited, frustrated, grateful, indifferent, involved, keen, on edge, optimistic, overwhelmed, panicky, positive, relieved, resentful, resigned, stressed, under pressure, worried

* The card sorting informed us about the strength of feelings and frequency of behaviors, allowing us to make conclusions about the third non-trust condition (ignorance). Based on: (Abrahms et al., 2003; Koeszegi, 2004; Lewicki et al., 1998; McKnight & Chervany, 2001; Saunders et al., 2014)

In order to understand the relationship between PhD students and their significant others over the course of the academic development program, we conducted 21 pre-program and 21 post-program interviews in English with all the PhD students who graduated from the program. We only interviewed PhD students with teaching experience prior to entering the program. Without such teaching duties, it is not reasonable to expect that PhD students would have formed teaching-related relationships or conversed about teaching when we first interviewed them. PhD students were at different stages of their studies. The most junior had just finished their first year when they entered the program, while the most senior were close to or right after dissertation defense.

The average duration of the 42 interviews was 28 minutes. All interviews were conducted in person (except for five – two pre-test and three post-test – that were done via Skype), and recorded. In the one case in which a technical issue made it impossible to use the recording, we relied on the interviewer’s notes. Each interview recording was transcribed verbatim and independently coded by two coders. In our analysis, we used a mixed method research design, combining statistical tests (cross tabulations, chi-square, and Cramer’s V) to investigate relationships between variables, with a qualitative content analysis of the interviews. During content analysis, coders first identified the information relating to each hypothesis in order to provide some empirical richness for findings. Coding differences were resolved through discussion.

Results

PhD dissertation supervisors were named most frequently, both before and after the program, as PhD students’ most significant conversation partners. Course leaders with whom participants co-taught courses were the second most common choices in the pre-program phase, while other faculty members and PhD supervisors were equally popular choices in the post-program phase. It is important to note that the two participants who could not name a conversation partner at their institutions before they enrolled in the program developed such a partner by the post-program phase (see Table 2). However, the degree of change is even greater than one would expect on the basis of Table 2, as almost half of the participants (N = 10) named a different significant other at the end of the year.

Table 2. The significant other relationships of PhD students before and after undergoing training

Pre-program		Post-program
Significant other	Frequency	Significant other	Frequency
PhD supervisor	7*	PhD supervisor	8*
Course leader	5**	Course leader	4***
Another faculty member	4***	Another faculty member	8*
PhD student	2	PhD student	1
No significant other	2	No significant other	0
MA dissertation supervisor	1	MA dissertation supervisor	0
TOTAL	21	TOTAL	21

* chosen by most participants

** second most popular choice

*** third most popular choice

Our research reveals that, for PhD students, trust is a central aspect of these relationships. Two-thirds of the examined relationships were trusting ones both in the pre- and the post-program phase (see Table 3). Participants’ appraisal of the significant other relationship also suggests that they find trust – or the lack thereof – important. They often described the relationship with reference to trust conditions even without being prompted to do so. For example, trusting participants talked of their relationships in the following terms: ‘it felt like, you know, we could trust each other, he had never done anything to … kind of betray me or I never felt betrayed by him’ (participant #18a, pre-program), and: ‘I trust him. Because I don’t feel intimidated by him and I’m quite sure when I’m around him, he’s not going to do anything to sabotage me’ (#6b, post-program). Conversely, a non-trusting relationship was described as one in which: ‘My supervisor and I we were not on good manners, and therefore I couldn’t trust her and I was … I had to be alert of what her plans are and regarding teaching I was also afraid that she will find it a problem that I’m teaching’ (#12b). Another talked of a similar relationship, saying that: ‘even in the communication between us … I had some distrust, you know? That he will not be on time’ (#17a).

Table 3. The distribution of significant other relationships across trust conditions

Pre-program		Post-program
Trust condition	Frequency	Trust condition	Frequency
Trust	13	Trust	14
No trust (ignorance, distrust, or ambivalence)	6	No trust (ignorance, distrust, or ambivalence)	7
No relationship*	2	No relationship	0
TOTAL	21	TOTAL	21

* participant could not name a significant other

Moreover, some participants found it equally important to be trusted. Five participants were conscious of the reciprocal nature of trust: of their significant other’s trust in them. ‘He had that trust in me’, one participant said. The same participant also saw independent teaching duties as a sign of trust: ‘He had the confidence … to allow me to try without having necessarily to sit at the back of the class’ (#6a). Another participant linked trust to encouragement: ‘I felt that, he trusts me, so, I had little problem with teaching of that course’. (#12a). However, in the pre-program phase, two participants talked of the significant other’s trusting in them as a burden rather than as a source of encouragement, feeling that such trust was undeserved. As one of them said: ‘I had this great support and I’m being trusted … I don’t want to lose this trust and I do not want to disappoint her. … Sometimes she sees me as she does herself, which might be [a] compliment for me … but I’m not [like her]’ (#5a).

Nonetheless, replacing one’s significant other appears not to be strongly related to trust. Only half of those in a non-trusting relationship had a new most significant teaching relationship by the end of the program, and the identity of five of thirteen trusted significant others also changed. When we look at the change in relationship from the point of view of trust, findings are more in line with our expectation: those in non-trusting relationships were more likely to have a new significant relationship by the end of the program. Whereas only one (16.6%) of the non-trusting relationships remained unchanged, eight trusting relationships (61.5%) did not change. However, although the association between trust at the beginning of the program and changes in trust conditions approaches statistical significance, it does not reach it mainly because we can observe the same number of changes (N = 5) from trusting to non-trusting as from non-trusting to trusting relationships (Chi² = 3.32; p = 0.069). That is, while 5 non-trusting relationships turned to trusting ones, an equal number of trusting relationships deteriorated into non-trusting ones. This supports our expectations that non-trustors would seek to improve their relationship in terms of trust. However, it also uncovers a disturbing trend regarding trust, raising doubts if, indeed, the pre-program nature of the relationship is the only driver of changes.

Participants’ explanation for changes in the relationship suggests that the presence of other – mitigating – contextual factors along with trust explains this trend. First, some noted the limited number of viable partners as significant others. Some relationships dissolved or changed in their nature, because the pre-program significant other left the institution (#17b, #28b) or the participants graduated from their PhD program and became faculty members (#18b, #19b). Proximity (for instance, sharing office space) and accessibility were also important factors facilitating change in selecting someone else for a conversation partner (#3b, #18b). Three cases suggest that issues in any of the trustworthiness dimensions – ability, benevolence, or integrity – are important reasons for participants to find a new significant other or question their trust in their significant other (#12b, #37b, #38b). Trust deteriorated in one relationship by the end of the program, because the significant other was not the person the participant had thought them to be (#36b). New significant others were sought out because the first significant other did not fulfill expectations of being an active teaching partner (#12b), and had a different – subject-focused instead of learning-centered – conception of teaching, not wishing to innovate their teaching (#37b).

Indeed, as the latter demonstrates, the program contributed to a small number of changes in relationships, which were also noted by other participants. For example, one of them said that because program requirements made them focus on the innovated course, the course leader became more important than their previous significant other (#7b). Finally, one person (#12) who had a distrustful relationship with their supervisor terminated the relationship by quitting their PhD program. This last example implies that certain significant other relationships can be consequential due to their troubled nature and that, while other significant other relationships can be formed in parallel, such a troubled relationship may have a negative effect on one’s academic career.

Trust clearly had a positive influence on how often conversations took place between each participant and their significant other in both the pre-program and post-program phases. Lack of trust had a significant negative effect (p ≤ .05) on the frequency of conversations, to the point that only those in a trusting relationship had weekly conversations (see Table 4). Frequent conversations tended to be periodic: discussions mostly took place during the teaching term. Cramer’s V reveals a moderately strong association between trust and the frequency of conversations, which again suggests that other factors may also be at play. For example, some trustors did not talk more often, because they themselves did not feel the need for additional conversations (#31a, #12a) or because their significant other was not available to meet more often (#8a). Unlike in the case of doctoral research, the responsibility for starting a conversation about teaching often falls to the PhD student in the relationship. Nonetheless, the fact that some non-trustors had some, albeit infrequent, teaching-related conversations seems to confirm our assumption that, in working relationships, interaction with non-trusted individuals cannot be entirely avoided.

Table 4. The impact of trust on the frequency of conversations before and after the program

		Weekly conversations	Infrequent conversations
Pre-program phase (N = 17, missing = 4) Chi^2 = 3.86, p = 0.049 Cramer’s V = 0.477	Trust	5	6
	No trust	0	6
Post-program phase (N = 21; missing = 0) Chi^2 = 5.25, p = 0.02 Cramer’s V = 0.500	Trust	7	7
	No trust	0	7

In cases where trust was missing from the relationship, participants did what they could to avoid certain types of conversations. We expected that the absence of trust would likely prevent conversations about methods, students, the syllabus, and teaching-related reflections, but not about content, assessment, and administrative issues, which we judged as unavoidable discussion topics. First, we checked if lack of trust resulted in zero conversations about avoidable topics and found our expectations confirmed in the pre-program-phase, but much less so in the post-program phase where lack of trust prevented discussions regarding syllabus, reflection, and, surprisingly, administrative issues, but did not prevent conversations regarding content and assessment and, unexpectedly, methods and students (see Table 5).

Table 5. The role of missing trust in preventing different types of conversations from taking place

	Expectations	Expectations confirmed?
		Pre-program	Post-program
Content	Some conversations	√	√
Methods	No conversations	√	X
Assessment	Some conversations	√	√
Students	No conversations	√	X
Syllabus	No conversations	√	√
Administrative issues	Some conversations	√	X
Reflection on teaching	No conversations	√	√

Second, we interpreted this hypothesis in terms of a relationship between two factors – trust and a particular type of conversation – and used chi-square and Cramer’s V tests to see if trust indeed had a positive effect on certain type of conversations (methods, students, syllabus, and reflections). Although all relationships pointed in the expected direction, the results are mixed (see Table 6). As expected, we found no significant relationship between trust conditions and what we deemed unavoidable conversations. However, a significant positive relationship between trust and discussions relating to students and syllabus design was confirmed only in the post-program phase. Reflections on the teaching process (including feedback and self-assessment) performed best: our expectations were confirmed in the pre-program phase and the significance level approached – but did not reach – statistical significance in the post-program phase. Taking into account all four tests regarding the relationship between trust and types of conversations, we found our expectations confirmed in 3 out of 4 cases regarding most types of conversations. The outliers were those relating to teaching methods, confirmed in only one case, and those relating to students, where only half of the tests fell in line with our expectations.

Table 6. The effect of trust on different types of conversations

		Confirmed?
	Expectations	Pre-program	Post-program
Content	No relationship	√ (2.72; 0.099; n/a)^a	√ (1.57; 0.210; n/a)
Methods	Positive relationship	X (0.93; 0.334; n/a)	X (2.42; 0.120; n/a)
Assessment	No relationship	√ (0.04; 0.852; n/a)	√ (0.33; 0.568; n/a)
Students	Positive relationship	X (2.24; 0.134; n/a)	√ (4.02; 0.045; 0.472)*
Syllabus	Positive relationship	X (0.93; 0.334; n/a)	√ (3.96; 0.047; 0.456)*
Administrative issues	No relationship	√ (2.36; 0.125; n/a)	√ (1.30; 0.253; n/a)
Reflection on teaching	Positive relationship	√ (5.60; 0.018; 0.632)*	X (2.96; 0.086; n/a)

^achi² value; chi² p value; Cramer’s V value

* statistically significant relationships

Finally, our investigation evidences positive change in significant other relationships with regard to the trustworthiness of PhD students, which overall documents a change in the role enactment of participants and in the role relationship with their significant others. To investigate this, we asked participants to assess changes in their role, relative power, and level of comfort in their relationships. Twelve of the 21 participants reported that their role in the relationship changed, while five reported no change in their role. Another five participants gave no straightforward answer regarding changes in their role. As for relative power, eleven participants reported that they had a more equal partnership (five reported no change, one reported a worsening in their relative power, and we had no information from five participants), making the relationship more favorable for trust. Concerning feeling of comfort, nine participants reported that, after completing the program, they felt more comfortable in their relationships (seven reported no change, two reported a decreased level of comfort; we had no information about three participants). Overall, 16 participants reported positive change in at least one of these dimensions (no or negative change = 3; missing data = 2).

Such positive changes were most often related to the enabling effect of the academic development program. Participants believed that the relationship improved because they felt more in control of their teaching (#5b, #22b), and more like a teacher and partner in teaching (#7b). They also felt able to offer something to the significant other, who in turn took it into account (#7b; #25b) and believed their knowledge and/or skills had grown (#5b, #8b, #13b, #8b). Participants also made the link between increased knowledge and the ability to converse about teaching and learning. One said: ‘the more knowledge you have, the more you can articulate it’ (#6b), while another noted: ‘I am more experienced, so I can … even talk about it more … . I can elaborate more on some things’ (#39b). Improved ability to articulate ideas about teaching and learning was also confirmed by the analysis of the transcripts from the interviews with program participants. Moreover, two participants noted that their significant others also recognized their increased pedagogical skills and knowledge (#8b, #22b). One of them (#8b) linked this to their significant other’s increased trust in them. One significant other was inspired by the pedagogical conversation with a participant and planned to use a similar method in their own course (#31b).

While participants’ trustworthiness increased regardless of whether trust existed between them and their significant other, trust – or the lack thereof – played an important part in influencing the behavior of participants in their significant other relationships. In general, the increase in trustworthiness was more likely to be recognized in trusting relationships, resulting in more equal partnerships and increased conversations and collaboration. In non-trusting relationships, empowerment was more likely to result in participants striking out on their own, making them less reliant on conversational input from a troubled relationship, which in turn decreased the importance of these relationships. A participant in a non-trusting relationship talked about gaining more independence by having the space to develop their own – different – teaching styles (#13b). Increased confidence in teaching-related knowledge and skills allowed another participant in a non-trusting relationship to judiciously ignore some of the demands of their significant other (#5b).

However, empowerment did not always influence relationships positively. One participant, who questioned their earlier trust in their significant other on the basis of his participation in departmental gossip and politicking, saw their increased knowledge and teaching skills as a foundation to further question their trust in the conversation partner: ‘I think he just do [sic] what he thinks is good [at] but he doesn’t take care about how to be better teacher, he takes care about how to lead faculty in better way but not about the teaching skills and being good as a teacher’ (#36b).

Conclusion and implications

Overall, our study leads us to conclude that the absence of trust seriously limits not only the frequency of pedagogical conversations, but also the diversity of issues discussed. At the same time, academic development programs can improve PhD students’ trustworthiness as they become expert conversation partners; however, they may also diminish their trust in significant others when, as a result of the program, a wide gap develops between the PhD students’ and their significant others’ conceptions of the teaching role. This study has also demonstrated that academic development programs may propel participants in non-trusting relationships to seek new significant other relationships.

While PhD students may have other important conversational partners, many of them named formalized relationships – dissertation supervisors and course leaders – as their most significant teaching-related relationships. This suggests that formalized hierarchical relationships are most likely to be seen as fostering important pedagogical conversations than looser relationships between individuals at the same academic position (for example, a PhD student with a fellow PhD student).

Based on the results presented in this article, we suggest that academic developers map the most significant teaching relationships of PhD students participating in an academic development program and direct PhD students’ attention to those relationships. Participants should be encouraged to interact more frequently with their trusted significant others, who could also be invited to become more integrally involved in academic development programs as natural allies and, potentially, advocates for the program. In addition, academic developers should encourage participants to look for trustworthy conversation partners whenever possible.

Acknowledgments

The authors thank Vicky Davies, Peter Felten, Katarina Mårtensson, Lynn McAlpine, Torgny Roxå, Agnes Simon, Mark Saunders, and two anonymous reviewers for their useful comments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was co-funded by the Erasmus+ program of the European Union (project No 2016-1-SK01-KA203-022551, ‘Extending and Reinforcing Good Practice in Teacher Development’).

Notes on contributors

Eszter Simon

Eszter Simon is aSenior  Lecturer in International Relations at Nottingham Trent University. Her research focuses on the role of trust in International Relations and academic development, and on the contribution of the Moscow-Washington hotline to crisis management. She has been co-investigator of several academic development grants and is a co-editor of Teacher Development in Higher Education: Existing Programs, Program Impact and Future Trends (2012, Routledge, with Gabriela Pleschová) and of the Handbook on Teaching and Learning in Political Science and International Relations (2014, Edward Elgar, with John Ishiyama and William J. Miller). She is aSenior Fellow of the Higher Education Academy.

Gabriela Pleschová

Gabriela Pleschová works at Comenius University in Bratislava, Slovakia, where she is the principal investigator of an international collaborative Erasmus+ project aimed at improving internationalisation through enhanced competences of university teachers. Her background is in Political Science, and in 2012 she earned an MSc in Education (Higher Education) from Oxford University. Since 2004, she has been coordinating workshops and other development activities for beginner university teachers. She serves as a member of the editorial board of the Journal of Political Science Education. She is the co-editor of Teacher Development in Higher Education: Existing Programs, Program Impact and Future Trends, published by Routledge in 2012 (with Eszter Simon). In 2019 she became a Principal Fellow of the Higher Education Academy.