How understanding Doctoral researchers’ coping strategies can inform Higher Education institutions’ response to their stress

ABSTRACT

Doctoral researchers report higher levels of stress and are at an elevated risk of mental ill health compared with undergraduate students and whole population normative data. Evaluating their reported coping strategies is essential for universities to develop interventions to address this. Content analysis was used to code 182 reported reactions to stress gathered via a qualitative survey. A framework based on a model of occupational stress in academic settings and additional inductively driven categories were used to code the data. It was also coded into maladaptive and adaptive responses. The most common adaptive strategies included ‘cognitive coping’, ‘social support’ and ‘active leisure’. Maladaptive strategies included ‘rumination’, ‘overworking’ and reports of having ‘no strategy’ to cope. Sub-group analyses showed some differentiation in coping strategies by gender and whether DRs were academic staff whilst frequencies of reported categories revealed different strategies were adopted at different stages of programmes. Based on these findings, the authors acknowledge that coping is a multi-dimensional process. The implications in terms of developing supportive interventions and mitigating less adaptive coping strategies are discussed.

KEYWORDS:

• Doctoral researchers
• coping strategies
• stress
• wellbeing

Introduction

There is a relatively small, but nevertheless significant corpus of research documenting the stressors experienced by doctoral researchers (referred to in this article as DRs). Whilst this population have historically been overlooked in favour of undergraduates (Watson and Turnpenny 2022), the publishing of some key papers, both in the UK and internationally, acknowledging that DRs are at an elevated risk of mental ill health compared with educated working professionals (Hazell et al. 2021), undergraduate students in higher education (Levecque et al. 2017), and whole population normative data (Hazell et al. 2020) has led to a stronger focus on DRs specifically. Importantly, this significantly higher prevalence amongst DRs could not be attributed to pre-existing mental health problems; instead, factors associated with doctoral research programmes have been evidenced to render DRs as vulnerable to stressors which precede mental ill-health (Hazell et al. 2021). Mackie and Bates (2019) suggest that stress levels may be exacerbated or in part attributable to factors unique to the environment of doctoral education (e.g. isolation, role conflict and supervisory relationships). It is essential that research in this area continues to enable an understanding of what contributes to high stress levels and to enable those with a responsibility for the wellbeing of doctoral cohorts to be informed about the bespoke challenges DRs might encounter. However, little attention has been paid to what DRs do in response to these stressors. We suggest that an interpretation of coping strategies adopted provides necessary information to better understand DRs behaviour in relation to stress.

Defined as ‘the thoughts and behaviours used to manage the internal and external demands of situations that are appraised as stressful’ (Folkman and Moskowitz 2004, 745), coping emerged as a distinct field of psychological inquiry during the 1970s and 1980s. Lazarus and Folkman’s (1984) transactional approach remains the cornerstone of research in this area, acknowledging that stress and coping involve a complex interplay between the individual and their environment. Considering the context of the perceived stress and how the individual appraises the stressors is central to understanding the coping strategies employed. A coping strategy is initiated in response to the appraisal. Individuals engage in primary and secondary appraisals which shape their emotional responses and determine the approaches they adopt to coping (Folkman and Moskowitz 2004). Primary appraisals allow consideration of the meaning of the potential stressor in a given context (e.g. might it be positive? benign? stressful?). If appraised as stressful, the outcome of the secondary appraisal is to acknowledge something needs to be done to address it (Biggs, Brough, and Drummond 2017). Taken together, these appraisals determine individuals’ reactions and behaviours when managing stress. In other words, if a situation is appraised as stressful and a threat (primary appraisal) the individual needs to evaluate their coping resources (secondary appraisal) and enact a coping strategy. A transactional approach views dealing with stress as a conscious, directed process.

Since the seminal work of Lazarus and Folkman (1984), coping has remained a central element of stress research, and numerous categorisation systems have been developed within the field. Although there is still no agreed consensus on how to measure coping (Biggs, Brough, and Drummond 2017), many context-specific measures have been validated, allowing researchers to focus on a distinct population of interest. The current research has applied categories from one such measure related to academia. Du Plessis and colleagues (Du Plessis 2020; Du Plessis and Martins 2019), proposed a set of strategies academics use to cope with the occupational stressors experienced. Their approach considers the role of appraisal in the enacting of coping strategies, with additional focus on whether strategies are adaptive or maladaptive and how well each predicts coping success. Following the construction and validation of a bespoke instrument to measure coping with stress in an academic setting (The Comprehensive Coping Strategies Questionnaire; Du Plessis and Martins 2019), a model comprising nine strategies was developed which indicated that adaptive strategies (cognitive coping, social support, and vacation time) predict coping success, whereas the opposite trend was observed for those strategies considered as maladaptive (avoidance, social disengagement, and rumination; see Du Plessis 2020).

This framework was selected for the current study because the academic environment is one where rates of occupational stress have been rising, with undertaking research often cited as a contributory factor (Levecque et al. 2017). DRs are embedded within this occupational environment, providing over half the research conducted by universities (Barry et al. 2018); studies often categorise doctoral students as members of ‘research’ or ‘university staff’ (Mackie and Bates 2019), thus it is likely that some publications describing occupational stress in academia will include data from DRs. Furthermore, as DRs work closely with up to three research supervisors or as part of wider research teams, their role within the academic environment arguably mirrors academic staff more closely than students on taught programmes, particularly when DRs support learning and teaching activities.

It is noteworthy, that some academic staff in the UK may also be enrolled on doctoral programmes (sometimes as a contractual requirement). This dual status can add further types of stressors such as increased workload (Smith et al. 2020); switching between identities, and feelings of vulnerability if supervised by colleagues (Billot et al. 2021). All of which might contribute to instigating different coping strategies to full-time DRs.

To our knowledge, no instrument for measuring coping strategies has been designed exclusively for DRs, therefore one validated within an occupational academic context represents the nearest ‘goodness of fit’ available. The current study sought to categorise DRs reported coping strategies using a framework based on occupational stress in academic settings, and then consider how the findings might be used to identify appropriate interventions to reduce stress amongst this population.

Materials and methods

Design

An online qualitative survey was designed to gather information about DRs perceived sources of stress and how they cope with this. Open-ended questions explored their accounts of stress as a doctoral researcher, and what they do when they experience stressors. It is the latter part of the survey, the reported coping strategies, which this paper reports.

To enable a more nuanced analysis, demographic information was also collected (gender, age, mode and programme of study, discipline area and stage of study); Participants were also asked to identify if they were a member of staff. Many academic staff in this sample were contractually obliged to undertake a doctorate, introducing the potential for the different pressures and coping strategies as discussed in the introduction; Hence, the decision to divide responses in this way.

The online approach allowed participants flexibility to take part in the study at their convenience. It was considered advantageous over individual interviews which could have potentially restricted sample size and reach (Braun et al. 2021). To preserve anonymity and confidentiality, all participants were asked to provide a unique identifier to be used in the event of requesting withdrawal of their data from the study. They were advised that there was no obligation to complete the survey or answer every question asked. Given the sensitive nature of the questions, links to internal and external support services were provided at the end of the survey. Ethical approval was granted by the University’s Ethics Committee.

Recruitment

This research took place in a UK university and was conducted across all five academic schools. Enrolled DRs at any stage of study were eligible to participate. DRs were contacted about participation through email, Twitter, and an advertisement on the University’s internal platform for DRs. Correspondence contained brief information about the purpose of the study and a link to the online ‘Qualtrics’ platform through which the survey was administered.

Analytic procedure

A qualitative content analysis was used to code and report the findings of this study, based on the analytic technique offered by Elo and Kyngas (2008). This involves three discrete phases (preparation, organisation and reporting) which, for transparency, are described in further detail below.

Preparation

Selecting the unit of analysis

Reported coping strategies were read several times by both researchers to initiate familiarisation with the entire dataset and ensure immersion in the data during the preparation phase (Elo and Kyngäs 2008). Having independently familiarised themselves with the responses, the researchers convened to discuss the unit of analysis, and approach to coding to be adopted.

The overall units of analysis were the response that participants provided to the question: What do you do when you experience such stress?

These ranged in length from a single sentence e.g. try to maintain a good routine. (P52)

To longer paragraphs e.g. Workload planning and prioritisation are relative strengths for me, so I can plan small, achievable targets for specific time frames. I also implement a number of self-care strategies, including regular yoga, meditation, walking and exercise. (P60)

Content analysis, as a flexible approach, is well suited to analyse multifaceted data sets (Elo and Kyngäs 2008). Responses often comprised multiple coping strategies, which required further refinement to ensure that unit of analysis captured all reported strategies. Each different strategy reported within a participant’s response was considered a single unit of analysis.

Organisation

A deductive analytic approach utilising the categories previously validated by Du Plessis (2020) was employed. These concepts were used to develop a categorisation matrix (Elo and Kyngäs 2008) from which data were coded.

The matrix was developed based on the original nine categories identified by Du Plessis (2020), with statements coded for exemplification of one of the categories. Table 1 illustrates definitions and examples of the nine dimensions used to code data in the current study and indicates whether each strategy is adaptive or maladaptive (see Du Plessis and Martins 2019 for all originally proposed dimensions and sub-dimensions). Each code was coloured either red or green to illustrate if it was an adaptive or maladaptive coping strategy. Following an independent coding process, the researchers compared their allocation of responses to the pre-determined codes identified in the matrix.

Table 1. Original coping strategies underpinning the deductive analytic framework (Du Plessis 2020).

Coping Strategy	Example	Type
Cognitive	I tried to replace negative thoughts with more positive ones.	Adaptive
Emotional	I allowed myself to express my emotions about the situation.	Adaptive
Social Support	I sought comfort from my social support network.	Adaptive
Active Leisure	I engaged in sporting activities such as playing golf, tennis, squash and soccer.	Adaptive
Vacation Leisure	I took a vacation.	Adaptive
Religious	I prayed to God.	Adaptive
Avoidant	I ignored the situation.	Maladaptive
Social Disengagement	I avoided contact with people.	Maladaptive
Rumination	I thought about the consequences of the situation instead of finding a solution	Maladaptive

To ensure completeness and saturation of the data (Elo et al. 2014), an inductive data-driven analysis was adopted for the second data organisation phase. Any responses which did not fit into Du Plessis categories were coded, leading to an additional categorisation framework drawing on the principles of an inductive approach (Elo and Kyngäs 2008).

Four additional categories were identified:

1. Reports of working very long hours were labelled ‘Overworking’ which the authors viewed as an unhealthy, maladaptive strategy.

2. Participants reporting not knowing how to cope or not having a strategy were labelled ‘No Strategy’ and considered maladaptive.

3. Several participants alluded to taking a break from their studies in response to stress; however, these DRs did not explicitly report this in terms of having a ‘vacation’ (a category in the original model). An additional leisure-specific category was introduced with terminology which better represented what was being described by participants; namely, ‘Other Leisure’ (i.e. strategies which involved taking a break which was not depicted as a holiday per se).

4. Multiple participants described non-active leisure coping in response to stressors (e.g. sedentary activities such as reading, listening to music and meditation). Largely described in a positive way, this was coded as an adaptive ‘Passive Leisure’ coping strategy.

The final number of categories used to code the responses was 13. Once all units of analysis were coded using these 13 categories, they were organised for presentation as a set of results in the final (reporting) phase of CA.

Results

Quantitative findings

Sixty-eight DRs completed the survey, offering 182 examples of coping strategies. This sample size falls in the mid-range for qualitative surveys (Braun et al. 2021). Demographic information is presented in Table 2. To allow for a more nuanced analysis, demographic data were further explored to identify variability in the responses provided by relevant sub-groups across the sample. This was achieved using SPSS v.24 (IBM Corp. 2016), and through splitting the dataset according to the sub-group of interest (i.e. student or staff) to enable comparisons, and identify differences between the coping strategies adopted.

Table 2. Participant demographic information.

Category	N	%
Gender Male Female Other Age (years) 21–25 26–30 31–35 36–40 41–45 46–50 51–55 56–60 61+ Mode Full-time Part-time Staff Yes No Stage (years) 1^st 2^nd 3^rd 4^th 5^th 6^th 7^th Not specified	24 43 1 7 13 13 9 7 7 8 3 1 29 39 30 38 9 18 20 10 4 3 1 3	35% 63% 2% 10% 19% 19% 13% 10% 10% 12% 5% 2% 43% 57% 44% 56% 13% 26% 29% 14% 6% 5% 2% 5%

Adaptive strategies accounted for 83.5% of reported coping strategies. Cognitive coping represented the largest proportion of strategies employed (30.2%), with avoidance the most frequently cited maladaptive stress response (6%). In total, 40 participants provided only adaptive coping examples, while nine utilised only maladaptive strategies, or no strategy at all. Seventeen participants adopted a combination of both adaptive and maladaptive coping in response to stressors. Table 3 provides a full breakdown of frequencies of each category.

Table 3. Sub-group differences in reported use of coping strategies.

Sub-group	N	Coping Strategy (M)
		Adaptive	Maladaptive
Gender Male Female Staff Yes No Stage (years) 1^st 2^nd 3^rd 4^th 5^th 6^th 7^th	24 43 30 38 9 18 20 10 4 3 1	75.47% 87.30% 82.13% 71.83% 94.44% 50.42% 88.16% 77.27% 62.5% 83.33% 100%	24.53% 12.70% 17.17% 28.17% 5.66% 59.58% 11.84% 22.73% 37.5% 67.77% 0

Sub-group analyses showed those with dual roles (staff/doctoral student) reported adaptive coping strategies more frequently than those who identified as full-time DRs (82% and 72%, respectively), but also opted to ‘overwork’ more often than non-staff (42% of maladaptive strategies provided by ‘staff’), whose predominant maladaptive response to stressors was ’avoidance’ (39% of maladaptive strategies provided by ‘non-staff’). Gender comparisons also revealed some variations. Females utilised significantly more adaptive coping strategies than males (MD = 0.89, p = .009). This was accounted for by significantly greater use of adaptive ‘social support’ coping (MD = 0.31, p = .003), and lower adoption of maladaptive coping through ‘rumination’ (MD = 0.14, p = .006), compared with male DRs. Differences according to stage of study were also observed with adaptive coping accounting for 50.% of strategies amongst second year DRs compared with 94% and 88% of first- and third-year DRs, respectively. This was accounted for by a greater proportion of second-year DRs providing examples of ‘avoidance’ (29%) and ‘rumination’ (24%) coping strategies specifically. A full breakdown of the strategies adopted by sub-groups is displayed in Table 4.

Table 4. Frequency of coping strategies adopted.

Coping Strategy (A/M)	Frequency (%)
Cognitive (A) Emotional (A) Social Support (A) Active Leisure (A) Vacation Time (A) Religious (A) Avoidant (M) Social Disengagement (M) Rumination (M) Other Leisure (A) Passive Leisure (A) Overworking (M) No Strategy	55 (30.22) 8 (4.40) 31 (17.03) 26 (14.28) 4 (2.20) 1 (0.55) 11 (6.04) 0 (0) 8 (4.40) 16 (8.79) 11 (6.04) 8 (4.40) 3 (1.65)

Qualitative findings

Categories are ordered from most to least frequently reported. F/M/NB indicates gender (female, male, non-binary) and S indicates staff member.

Adaptive strategies

Cognitive coping

Cognitive coping strategies often involved practical ways to organise work tasks, with participants describing writing a plan or a ‘to do’ list. P22 (F) described how she would, ‘ … draw up achievable writing charts’. The word achievable indicates such strategies can facilitate productivity yet maintain a sense of perspective about what can be done. Lists also offered reward and motivation:

P15 (F):‘I still set myself goals and weekly “To Do” lists. Lists are probably the most useful thing for getting myself to complete jobs as I enjoy the experience of ticking things off’

The use of adaptive cognitive strategies manifested in a variety of forms which may contribute to wellbeing and academic progression through reducing ‘self-sabotaging’ behaviours (i.e. procrastination and disorganisation) which can often impede DRs from attaining their goals (Kearns, Gardiner, and Marshall 2008). As demonstrated in Kearns, Gardiner and Marshall’s 2008 evaluation of a programme for ‘PhD success’, establishing a clear, specific goal before planning how and when this may be accomplished can reduce DRs’ stress and improve completion rates and durations. Though on occasion, such coping strategies perpetuated further stress and anxiety:

P63 (F): ‘I cope with worries about timeline issues by regularly revisiting Gantt charts and mapping progress. This does help somewhat but also causes more concern if I can see the project slipping further behind’

Social support coping

As the next two extracts illustrate, simply talking to others played an important role when experiencing stress:

P64 (F, S): ‘Even though it is the last thing I want to do sometimes, talking to everyone and anyone who will listen about managing projects or a specific problem I’m facing always helps’

Respondents reported speaking to a range of individuals both internal and external to the university and the doctoral process. Those external were predominantly family members or friends:

P15 (F): ‘It is also about communicating with family, friends and my partner when I am struggling. Talking helps a great deal’

Support from within the university included other staff members or DRs, whose function was to give advice, rather than generic comfort or escape:

P76 (F): ‘Talk to others who have also experienced the doctoral journey to find out how they managed to overcome these issues’

The role supervisors played was more directive:

P37 (NB): ‘Arrange a meeting with my director of studies who helps me work out what I should be focusing on and helps me break work down into manageable chunks.’

Previous research indicates that the effectiveness of support is largely dependent on whether the source of stress is academic or emotional (Byers et al. 2014; El-Ghoroury et al. 2012). The current study similarly indicates that social support is not unidimensional, with DRs seeking support from those who fit their specific need (e.g. supervisors when requiring academic guidance; peers for companionship; and family when seeking emotional support).

Active leisure coping

Active leisure was the most frequently reported leisure coping strategy, comprising multiple activities:

P18 (F, S): ‘Go for a run’

P47 (M, S): ‘Exercise (walk, swim, yoga, run)’

Engaging in physical activity is a well-documented way to enhance wellbeing/mood and relieve stress (e.g. Marquez et al. 2020). This is corroborated by participants who explicitly reported the benefits and effectiveness of this strategy:

P5 (F): ‘I still make sure that I take my daily walk first thing every morning because the benefit to my wellbeing is immeasurable’

Engaging in physical activity or exercise is also commonly reported as an effective response to stress amongst DRs (El-Ghoroury et al. 2012; McCray and Joseph-Richard 2020). Our participants may have accrued further benefits from their ‘active leisure coping’ taking place in an outdoor environment. This has been endorsed in extant literature evidencing a strong positive association between nature-based recreation and indicators of mental health (Lackey et al. 2021), and green space and stress relief (Van den Berg et al. 2010).

However, such strategies require regular engagement for long-term benefits, as acknowledged by the following participant:

P67 (M): ‘Exercise relieves stress, temporarily … ’

Additionally, taking time out to exercise can itself become a source of stress:

P8 (F, S): ‘Usually I turn to exercise to relieve the pressure but then feel guilty that I have not been working’

Passive leisure coping

Passive leisure provided an opportunity to engage in something different as a temporary distraction from the research degree, for example:

P53 (F,S): ‘Read a book, colouring, listening to music’

Meditation was also coded as passive leisure for the purposes of relaxation and/or stress reduction. This popular activity represented an example of a range of ‘self-care strategies’ (P60, F) adopted. Passive leisure coping by means of meditation may also be utilised to mitigate the potentially detrimental impact of an emotional coping response to stress, through controlling and re-framing those emotions experienced:

P73 (F) ‘Whenever I feel stress associated with the above stressors, firstly I worry and panic. But later, practice mindfulness techniques which allows me to better rationalise my negative thoughts associated with these stressors’

Such accounts reinforce findings for the effectiveness of mindful meditation on doctoral researchers’ psychological wellbeing, and further illustrate the association between mediation and self-efficacy through an increased sense of control over thoughts and emotions (Barry et al. 2018).

Other leisure coping

This category was added to account for activities which did not fit into active or passive leisure, or vacation time, but nevertheless allowed participants to take an interlude from their academic work. For example:

P85 (F): ‘Having a cup of green tea’

P53 (F, S): ‘Treat myself to a nice coffee’

This illustrates that such activities are pleasurable means to take a break from the work. Some participants did not cite the nature of the activity but described taking time out from studies:

P29 (M): ‘I manage my time and move away from the problem. Coming back to task after a break from it, really helps to focus the mind’

Others explained how an initially adaptive strategy like taking a break can, ironically, engender a maladaptive response (here ‘overworking’) to mitigate the impact of this ‘non-working’ time:

P80 (M): ‘Stop what I’m doing and have a few hours to myself before starting again or I tend to work more than I should to cancel out the feeling of not succeeding to be productive’

Vacation time

Only one participant referred explicitly to travelling as a means of relieving academic stress:

P91 (M): ‘Depended on the amount of responsibility I had. The first year, I travelled a lot’

This indicates that P91, now in their final year, had more leisure time during the earlier stages, and reinforces our findings that coping strategies adopted can vary according to stage of study. Other responses in this category related to short-term blocks of time spent away from the PhD-related demands:

P55 (F,S): ‘try to block out periods where I know I can work on the PhD and periods when I don’t/can’t – a little PhD holiday’

The duration of ‘vacation time’ as an effective strategy for coping with academic stress was highly variable between DRs. This could be indicative of the amount of perceived stress and fatigue experienced at the time thus impacting on the optimal ‘recovery time’ required (Blasche, Zilic, and Frischenschlager 2016).

Emotional coping

Whilst the Du Plessis model categorised emotional coping as an adaptive response to academic stress, narratives from participants in the current study suggest otherwise. For example:

P11 (M, S): ‘I get incredibly frustrated and angry with myself. Usually I turn to exercise to relieve the pressure but then feel guilty that I have not been working’

In contrast to the use of ‘social support coping’ it was evident that some participants’ emotional response to stress can negatively impact on their family members, and potentially be detrimental to academic performance and progression:

P85 (F, S): ‘ … shout at my family, get cross and sad about things, lose motivation’

P15 (F, S): ‘Shout at my kids more!’

Such accounts introduce an element of doubt towards the efficacy of emotional strategies as ‘adaptive’ and may partly explain why the original model found no association between emotional coping and coping success (see Du Plessis 2020).

Maladaptive strategies

Avoidant coping

Some participants cited ‘avoidance’ as a coping response to stress:

P1 (F): ‘Avoid the issue!’

P90 (F): ‘Avoid all responsibility and work’

Others illustrated why this was inversely related to coping success:

P37 (NB) ‘Ignore it until it blows up in my face, usually’

DRs avoidance strategies can involve not addressing academic tasks (see Hazell et al. 2020), as described by P54 (F):

Cannot concentrate and focus on what needs to be done academically and have put off completing things when I know that I need to get on with this

Some participants reported employing forms of behaviour which could be considered detrimental to wellbeing. This is poignantly illustrated here:

P42 (M, S): I don’t sleep, I drink in the evenings to relax and then a nightcap to get me to sleep and when I wake up in the night I drink to get back to sleep again

P9 (M, S): ‘I drink alcohol most evenings to help calm me down’

The need to ‘calm down’ suggests that there is some raised emotion, which again leads us to contend that the emotional coping strategies reported in this study are not an adaptive way to cope.

Rumination

Often related to the ‘emotional processing’ construct of emotional coping (Moreno, Wiley, and Stanton 2017), reflecting on, and exploring perceived academic stressors appeared to predict rumination for several participants in the current study:

P16 (F): ‘I sometimes spend hours staring at my computer almost paralysed with fear unable to start. Once I do manage to get going this subsides’

This use of the word ‘paralysed’ conjures up a powerful image of the debilitating impact rumination can have on academic progression, and mirrors findings which attribute DRs’ loss of self-regulation to feeling overwhelmed by the demands of doctoral research (Hazell et al. 2020).

Overworking

‘Overworking’ was considered maladaptive as working more hours does not increase the reported quality of work. Conversely, it often predicts an increase in stress by augmenting perceptions of work overload (Abbas and Roger 2013):

P12 (F, S): ‘If I am experiencing overwhelming stress then I often find it hard to focus and experience a decline in wellbeing, I often end up working long hours and producing work I am not happy with’

Unhealthy patterns of working hours were commonly reported within this category:

P41 (F) ‘I try to switch off from work before dinner/around 8 pm (but this doesn’t always happen – I’m completing this survey at 9:30pm!)’

There were no positive responses associated with overworking, an approach which was more commonly adopted by staff members, possibly because they could see no alternative when juggling work and study.

No strategy

Three participants provided accounts which indicated an absence of coping strategies in response to academic stress, for example:

P14 (M): ‘I have no coping strategies that I have identified as effective for me personally’

P36 (M): ‘I have had a particularly bad time with stress and depression which has left me feeling “burnt out”. I experienced a sustained period where I was unable to study or deal with work outside of research’

As depicted here, a lack of coping resources is frequently related to psychological distress and can precede common psychological disorders such as anxiety and depression (Taylor and Stanton 2007).

Discussion

This study has identified a range of coping strategies reported by DRs in response to stress. The approach taken makes it unique in the context of DRs, and provides a novel, and timely addition to the extant literature on stress amongst this population. As McCauley and Hinojosa (2020) acknowledge, many models of occupational stress exist but these are rarely used to consider the experiences of DRs. The rationale for using a model of occupational stress which was validated in an academic environment was offered in the introduction, within this discussion we critically consider the findings further and how they might inform practices adopted by those responsible for the progression and pastoral care of DRs in universities.

Analysis indicated that DRs adopt predominantly adaptive coping strategies, citing ‘cognitive’, ‘social support’, and ‘active leisure’ coping as the most frequently utilised strategies, respectively. This supports the model of stress and coping developed by Du Plessis and Martins (2019), which found cognitive and social support-related strategies to be a salient method adopted by academics irrespective of contextual variation. This is encouraging, given that these strategies are most likely to predict coping success (Du Plessis 2020).

‘Avoidance’ was the most frequently reported maladaptive coping response in the current study. While research indicates that avoidance-oriented coping can be effective for short-term, uncontrollable stress (Taylor and Stanton 2007), for DRs it is plausible that this approach may create a deleterious cycle between procrastination, low motivation, reduced enjoyment, and thus further avoidance of the perceived academic stressor (Hazell et al. 2020). The overall prevalence of avoidance as a coping strategy was relatively low (6%), however it is important to consider the potential impact of such strategies on progression, especially if it leads to ‘rumination.’

Participants rarely described just one strategy, and avoidance was often followed with an example of the replacement behaviour which, at times, was adaptive (e.g. talk to friends), but at others maladaptive (e.g. consume alcohol). This suggests that some caution needs to be exercised when discussing ‘adaptive’ or ‘maladaptive’ categorisations as binary, mutually exclusive categories. Similarly, one should refrain from considering any coping strategy identified in isolation. Not only might finer contextual differences be lost, but this could lead to conclusions based on only part of the whole. It was evident that coping when undertaking doctoral programmes is a complex, multidimensional process. To understand this further, we acknowledge that multiple strategies may be more likely to precede coping success and reduce perceived stress. Likewise, a range of supportive interventions are required at institutional level.

Context-dependent variations

The absence of ‘Social Disengagement’ as a (maladaptive) strategy to cope with academic stress may reflect contextual variation between the present study and research undertaken to validate the original model (Du Plessis 2020; Du Plessis and Martins 2019). Given the timing of survey completion (during a global pandemic), it could be that enforced social disengagement by virtue of pandemic-related restrictions may have become ‘normalised’ at that time, and thus not identified by participants as a strategy per se. It is also plausible that DRs who described ‘overworking’ to cope with stress may have indirectly disengaged from social networks to accommodate this transference of work-life balance.

The other category where there was a noticeable absence was ‘Religious Coping’, which accounted for only a single response. This may reflect differences between the cultural contexts between a largely religious (South Africa- setting for Du Plessis’s research) and secular (UK- setting of the current study) society. Previous research with university students has identified cross-cultural variation in the prevalence of religious coping in response to stressors (Chai et al. 2012) including a global comparative study which identified South Africa as one of only three countries worldwide in which religiosity is positively associated with health (Zimmer et al. 2019).

Staff

Sub-group analysis revealed that DRs who ‘overworked’ were more likely to be staff, with those who identified as ‘students’ predominantly adopting ‘avoidance’ strategies to (maladaptively) cope with academic stress. While such ‘self-sabotaging’ behaviours are commonly reported amongst DRs (Kearns, Gardiner, and Marshall 2008), this may also account for the absence of ‘social disengagement’ in the current study, given that students’ social networks are more likely to be their peers compared with academic staff (El-Ghoroury et al. 2012). It was also apparent that DRs who are staff predominantly employ problem-focussed coping (e.g. proactively approaching the stressor, albeit through unhealthy working practices), while students who have a greater tendency to employ emotion-focused strategies (e.g. avoiding the stressor and disengaging from the academic environment) (Abbas and Roger 2013). It is possible that the prior experience of academic stress experienced by staff may have increased perceptions of control towards stressors (Thompson 2002), with awareness of available resources and tools enabling an active rather than passive coping approach (Clouder et al. 2020). These differences support related literature highlighting the need for tailored support for academic staff undertaking doctoral studies, such as facilitating work-life balance and providing bespoke supervisory training (Billot et al. 2021).

Gender

Sub-group analysis revealed gender differences with females reporting significantly more adaptive coping strategies. These were accounted for by greater use of social support, and significantly less rumination than male DRs. This may be attributed to gender differences when utilising an emotional processing (EP) approach to coping. Previous research indicates that for females, EP is largely adaptive and includes exploration of emotions often via social support networks. For males, EP is not an adaptive strategy and indeed correlates positively with greater rumination (Moreno, Wiley, and Stanton 2017; Stanton et al. 2000). Despite these differences, these findings suggest that interventions such as DR-specific ‘writing retreats’ (which by their very nature incorporate time management and encourage socialisation), could be effective for increasing DRs’ productivity and progression irrespective of gender or associated coping strategies (Vincent et al. 2021).

Stage of study

There were also variations in coping strategies according to DRs’ stage of study, with 95% of strategies in year one deemed to be adaptive, compared with just 50% in year two. Further analysis revealed that DRs display greater rumination and avoidance towards their studies during the second year, which is acknowledged as a key ‘pinch point’ within the wider literature (Lane et al. 2019). Indeed, there is evidence to suggest that increased pressures and demands of the project, perceived lack of resources required, and perceived absence of appropriate social support can culminate in procrastination, reduced motivation and increased levels of stress and anxiety of DRs during this stage (Barry et al. 2018). Additionally, the adaptive benefits associated with ‘active leisure’ coping in year one may be perceived as avoidance of studies in year two, due to the increased demands DRs experience during this stage. It is suggested that a range of interventions are adopted to support DRs with some being focused on specific stages of the programme. Lane et al. (2019) suggest that mentoring may be important in the early stages, resilience training during mid-stage and writing clubs to support those working on correcting in the final stages.

Effectiveness of coping strategies

The current study did not require participants to comment on how effective reported strategies were for alleviating stressors associated with doctoral research. Nonetheless, the detail of participants’ responses afforded some insight into the effectiveness or outcomes associated with specific coping strategies. In conjunction with wider literature on stress and coping, this informed the following suppositions on the degree to which strategies precede ‘coping success’.

Findings illustrated that whether strategies are deemed to be ‘adaptive’ or ‘maladaptive’ is not always indicative of subsequent ‘success’ or ‘failure’ in coping with the stressor. For one DR, ‘active leisure’ may help to alleviate anxieties and improve cognitive functioning, whereas for another the same strategy may lead to guilt about spending time away from their studies. While cognitive coping was most salient, and largely described as adaptive, there was evidence that the specific technique adopted (e.g. Gantt charts) can determine effectiveness and even exacerbate stress, should the DR experience difficulties in adherence to milestones or achieving tasks. Workshops on time management and project milestones need to support DRs to manage realistic expectations, prioritise tasks and acknowledge research does not always go to plan.

Social support was identified as an important resource to mitigate and prevent potential stressors, with effectiveness largely determined by goodness of fit, rather than the amount of support available to DRs per se. From the responses provided, social support for DRs may be derived through various sources including supervisors, peers, friends and family. Yet the utility of each is likely to have varied, particularly in situations where the DR’s support requirement mismatched the type of support offered by the provider (Sufyan and Ghouri 2020). Typically, DRs can experience numerous challenges throughout their doctoral journey, which broadly emanate from either academic or personal/emotional stressors (Barry et al. 2018). When the stressor is academically related support from supervisors and peers has been most frequently attributed to success by DRs who have completed their studies (Mantai and Dowling 2015). Conversely, support from family and friends is most frequently sought, and often most effective when the source of stress is emotional or personal (El-Ghoroury et al. 2012). Given that both emotional and academic stressors are reported amongst DRs throughout their candidature (Barry et al. 2018), it is imperative that institutions identify those who may be most ‘at-risk’ of insufficient support networks (e.g. international students), to ensure these DRs have access to appropriate support (Waight and Giordano 2018).

Conclusions

By identifying how DRs cope with academic stress we can begin to implement systemic changes on an individual, organisational, and policy-level. Coping is multi-dimensional and a complex phenomenon, those who are responsible for the progression and pastoral care of DRs must acknowledge this and have a range of strategies in place to support and to assist. This study found cognitive coping strategies to be the most frequently utilised by DRs when experiencing academic stressors. Consequently, workshops, training, and resources which provide DRs with suggestions and methods of how cognitive coping strategies may be employed, could reduce or even prevent the stress associated with doctoral research. The mechanisms through which such interventions may be effective include increased self-efficacy following writing retreats (Vincent et al. 2021); improved self-management through time management skills (Kearns, Gardiner, and Marshall 2008); and increased resilience through self-managed psychoeducational programmes (Bekki et al. 2013). As a minimum, universities have a responsibility to offer DRs cognitive-based support provisions to develop their self-regulatory skills (Ryan, Baik, and Larcombe 2022), and, ideally, offer task-specific support (e.g. academic writing) which reflect prominent stressors encountered as DRs progress through their doctoral journey (Barry et al. 2018). We would like to see universities adopting such strength-based approaches which have the potential to shift the rhetoric from ‘surviving’ a doctorate to ‘thriving’ as a doctoral researcher.

Findings presented in this paper also suggest that there are means through which universities could indirectly promote the effectiveness of coping strategies adopted by their DRs. For example, key ‘actors’ such as DR supervisors should be endorsing and indeed modelling time away from studies (through Leisure coping or Vacation time), as a means of reducing feelings such as guilt which often negate the benefits of such coping strategies. In accordance with findings published elsewhere, unhealthy working practices have become the ‘norm’ amongst researchers in the UK (Moran et al. 2020), indicating that such practices may necessitate policy-level changes which promote a healthier research culture for DRs.

Disclosure statement

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