Table 2

Case studies in preparing for remote data collection

PrincipleCase studies on conducting remote data collection on violence during COVID-19
(1) Designing remote data collection in the context of strong research partnershipsMaisha Fiti study, Kenya
Research conducted with female sex workers in Kenya faced challenges, as many women lived with children, sometimes partners or other family members, and experienced insecure housing and work patterns due to the pandemic. At the same time, however, we also noted that working with sex workers could be different to other populations of women as they more often live without partners. We therefore felt that the risks were lower than with studies with cohabiting or married women, for example. Furthermore, we also noted some benefits of a flexible and remote approach with participants, as the study team was able to reach participants who would otherwise have been unable to participate due to migration and unstable housing, caused by economic difficulties of the pandemic. Here, we note how working with specific populations may offer certain challenges, but also particular opportunities for engagement. Remote methods research on violence highlights the impossibility of a one-size-fits-all approach and emphasises the need for participant-specific planning.


CCSS-Z study, Zimbabwe
In the CCSS-Z study, the research team did not feel the relationships with school staff and parents were in place to conduct remote interviews about violence. At the time of data collection in Zimbabwe, the government had recently signed into law Education Act Amendment No. 15 (2019) prohibiting corporal punishment in schools. This period of change had led to a sense of insecurity and sensitivity around this topic, and an increased media focus, with some undercover media reporting on violence using photos and video footage being published and shared on social media. This had heightened a generalised sense of distrust in and around schools of researchers, particularly for those making approaches over the phone, and a sensitive social and political context to be asking questions about violence in schools. Our researchers were highly cognisant of this and advised adapting the study design to avoid remote interviews with teachers and to place great emphasis on trusted pathways and partners through which we approached participants. Researchers waited until it was safe to conduct interviews with school staff and parents in person. Researcher knowledge of the context was essential, as well as a built-in capacity for the study to be adaptive and responsive to contextual needs of the moment.


Natsal study, Britain
The Natsal study is led by a multidisciplinary team of researchers across academic institutions and a national social research organisation. In addition, Natsal has a number of collaborators to support the topics addressed in the survey. Input into the questions and question wording for the sexual violence module in the fourth round of Natsal was obtained from researchers, practitioners and survivors’ organisations working in this field. The study was able to draw on the expertise and knowledge of survey methodologists to adapt the study design and data collection protocols for remote methods. The team considered and assessed a range of remote data collection models in terms of their ability to deliver the key design features of the Natsal study (i.e. probability sampling, minimising response bias, questionnaire length and sensitivity, a combination of interview and self-completion questions, biological sampling and maintaining a 30-year time-series). A recommendation was made to implement an interviewer-administered approach where initial contact was made in-person, face-to-face data collection offered with alternative remote modes available (video or telephone interviews with an online self-completion questionnaire). The team made significant adaptations to various aspects of planned study: fieldwork documents were modified, the interview instrument was adapted for remote modes, the self-completion was converted into an online questionnaire, remote biological sampling protocols were developed, paper consent forms (for biological samples and data linkage) were converted into eConsents, remote incentive administration was established and researcher training was adapted for online delivery.
(2) Safety and privacy of remote methodsBantwana programme study, Uganda
In our original research plan, in addition to interviewing caregivers and a range of adult stakeholders, we intended to conduct in-person focus groups with children in school settings. In response to the COVID-19 pandemic, we altered our research design and decided to conduct remote interviews with caregivers and adult stakeholders as we considered that remote data collection with children was not appropriate. This was primarily due to safeguarding concerns and practical constraints, such as how to safely reach children without going through schools.


Natsal study, Britain
In the Natsal survey, if a participant selected that they ‘prefer not to answer’ for any of the questions on violence, a question appeared asking whether they wanted to skip to the end of the section. Within the violence modules, there was also a button that the participant could click to take them to a neutral news website if there were issues with privacy. Throughout the online questionnaire, on completion of a section of questions, each section was locked to prevent the possibility of anyone going back to view previous answers. All participants taking part in remote method data collection were provided with an interview document pack containing participant facing materials (e.g. the participant information leaflet, interview show cards, a signposting leaflet).


Maisha Fiti study, Kenya
Prior to the main interview, the study team designed a process to have initial phone calls with participants to understand if they were interested in participating and to assess safety in their current location. Study procedures were also explained, and a time and day for the interview was agreed on. Technological challenges were also assessed and researchers discussed the following with participants: charging their phones, phone audibility, locating network friendly points, finding a safe place where people would not eavesdrop, and agreeing a safe word to terminate the interviews.
(3) Safeguarding processes in the context of remote data collectionCoVAC study, Uganda
As part of the process of moving to remote methods, we engaged in a substantial revision to our directory of referral organisations. We contacted all referral organisations that we had previously worked with to find out if they were still operating during the pandemic period and if they could take up ourreferrals. As several had closed or we had lost contact with some of our contact persons, we added further organisations to our directory. Local organisations were prioritised to hasten the referral process in case of severe cases that required immediate in-person engagement. Given the effects of the lockdown on the economic situation of participants, we also sought local organisations that could provide social services or support with income generation. All local organisations were contacted by the study counsellor, and formal letters were sent as a follow-up to share information about the study. In the quantitative study—where we asked questions about violence—we offered tele-counselling to every participant. In survey programming, we embedded algorithms in the questionnaire to help to identify which participant needed what type of referral intervention. All participants that met any of the referral criteria and accepted counselling were assigned to study counsellors for phone counselling. In the qualitative study, telecounselling was offered to those participants who reported violence and were willing to talk to the counsellor. All telecounselling was done by full-time study counsellors and any other referrals that needed in-person engagement were coordinated by the study’s lead counsellor.


CCSS-Z study, Zimbabwe
In in-person interviews in a Catholic school, one headteacher disclosed knowledge of violence against a child in his school. This met the criteria for referral within our mechanisms; however, the headteacher assured the researcher that they were handling the case internally within the school’s existing referral mechanisms and requested that it was not referred to our partner organisation, Childline Zimbabwe. This left the researcher in a difficult position as she felt uncomfortable forcing a referral to our partner organisation and did not want to over-ride the headteacher’s practice within the school and threaten relationships. Furthermore, the school had robust and comprehensive child protection mechanisms in place due to a prior child protection programme being conducted. At the same time, she had concern for the child and the case warranted referral based on the referral protocol. The study team met with our referral partner organisation, Childline Zimbabwe to discuss this case. We determined a course of action to support the researcher and amended our referral protocol to reflect the particularities of this case, in case something similar arose in the future. While this case occurred during face-to-face research, the ethical quandaries it posed were heightened by the researcher connecting with the study team remotely, feeling more isolated than would otherwise be the case, in dealing with this tricky decision. This highlighted the importance of remote support for researchers conducting violence research remotely and highly detailed referral mechanisms.
(4) Remote support and training for researchersHERA study, Nepal and Brazil
The HERA study developed a researcher distress protocol that partners adapted to the local context. The protocol described the processes for responding to and referring women who experienced distress during interviews, or who disclose traumatic experiences for which they wish to access psychological support. The protocol provided examples of mild, moderate and major distress symptoms and potential responses. It made clear that the researchers should not act as a counsellor, but rather offer reassurance, empathetic listening and referrals to appropriate sources of support. In terms of researcher safety and well-being, the protocol outlined a number of steps to minimise risk when undertaking interviews (e.g. sharing details of field visits in advance, working in pairs, using a checking-in system). It also recommended debriefing and support for researchers (e.g. regular team debriefing meetings with the PI to discuss emotional aspects of the work and address any particular issues that have arisen, or to ensure access to psychological support if needed). In addition, study teams in Nepal and Brazil had daily debriefs and expanded researcher training to include training on remote data collection. Data packages were also purchased for all researchers.


CoVAC study, Uganda
In anticipation of an increase in COVID-19 cases, we amended our study protocol to include measures to facilitate remote training and data collection to ensure safety of both the researchers and participants. These measures included carrying out an assessment with researchers to understand the feasibility of home-based data collection, providing researchers with additional training on data management and storage, and purchasing opaque folders and headphones for all researchers. We also ensured all study tablets and mobile phones were password protected and encrypted, developed a code of conduct for home-based data collection, and conducted regular well-being debriefs with researchers.


Natsal study, Britain
The Natsal study included both in-person and remote data collection options. The study introduced COVID-19 fieldwork protocols and training to ensure the safety of participants and researchers during doorstep contact and face-to-face interviewing. The guidance included regular personal health checks for researchers, social distancing and hygiene measures, PPE provision, materials and equipment handling and participant COVID-19 screening. The researchers also attended video interview training that covered the protocols for setting up and administering interviews using Microsoft Teams.