Ethical considerations

Several ethical issues must be considered when conducting diary' research (Filep, Turner, Eidse, Thompson-Fawcett & Fitzsimons, 2017). In particular to a mixed methods design, important considerations in our study were issues related to privacy and anonymisation of participants’ responses. As we were taking an in-depth look at doctoral students’ experiences within single departments, the participants were well known to one another and to other non-participants in the department (including their supervisors or other staff), which increased risks that participants may be identifiable in the dissemination of our findings. This was exacerbated in a mixed methods design, as triangulating multiple pieces of sensitive information about individuals could make it easier to ‘guess’ who the individual might be. For instance, we explored demographic and lifestyle variables related to students’ experiences, and there were instances when there was only one individual who matched a certain profile in the department. In these cases, we needed to take consistent and carefi.il consideration in writing up and disseminating our findings to ensure that the way we discussed participants’ experiences respected their privacy'. In some cases, this meant choosing not to share certain results or excluding certain examples in our writing, as the explanation would compromise a particular participant’s confidentiality.

As our participants discussed their daily' lives in the diaries and provided locationspecific details throughout our study, our data collection included many identifying markers, such as discussions about other doctoral students, supervisors, staff members, and university' or departmental signifiers. For example, there were instances in which participants spoke about negative impacts their doctoral supervisor had on their social support provisions, where that supervisor was named or described. We recognised that this had implications on several levels, should carefi.il steps not be taken to ensure anonymisation. Further, we needed to consider the privacy and dignity' of non-participants, who might find themselves identifiable or recognisable in the dissemination of our findings. We also felt there were potential issues related to power dynamics and social harmony' should a supervisor or staff member recognise negative sentiments expressed by their student, particularly considering we were researching within our own departments (Malone, 2003). Therefore, the project required a strict anonymisation protocol that was developed at the start of the study and was revisited throughout the data collection and analysis processes. This included removing names and place markers and deleting any statements that may have made the participant identifiable. We also made sure that all transcripts were anonymised by' members of the research team who were not in a position of power over the participant (such as being their supervisor). At the same time, the mixed methods design required the researchers to have a signifying marker (for us, a unique identifying number) to link the data from multiple collection points. An additional step was taken by sending participants a copy of their anonymised data and allowing them to make any additional anonymisations if they' felt uncomfortable or unsure whether they' could be identified, of which several participants suggested further changes.

Our research focused on a highly personal and sensitive topic and we recognised that social transitions can have an influential impact on higher education students’ mental health and well-being. Given that we collected such in-depth data about challenging subjects such as social isolation or social anxiety, we also recognised the potential that our participants may disclose information that might lead the research team to be concerned for their well-being or safety (for example, suicidal thoughts or severe anxiety). This is a particular concern for diary studies, given they often facilitate in-depth and personal reflection on sensitive topics (Harvey, 2011). As such, the research team developed a protocol at the start of the study for addressing any issues that might arise, including outlining what circumstances might warrant intervention, dedicating who would be designated with addressing any concerns and establishing easily accessible resources and referrals for students in crisis. However, it is worth noting that, unlike interviews, diary data are often submitted at particular intervals and may not be immediately available to the researcher, leaving no recourse for immediate action to safeguard participants.

Altogether, it is important to be aware that the nature of mixed methods diary research means more complex, personal and identifiable information is often collected about an individual, giving way to an increased potential for ethical issues. Developing clear protocol at the start of a mixed methods diary study for issues such as privacy, anonymisation, opportunities to withdraw data and circumstances that warrant breaking participants’ confidentiality is essential for supporting ethical research and easing the concerns of both participants and researchers alike.