We only use your email address to send you the newsletter and to see how many people are opening our emails. A full privacy policy can be viewed here. You can change your mind at any time and update your preferences or unsubscribe.

Data collection in covid-19 restrictions

An in-person interview following resumption of fieldwork. Photo: Mindset

Jordan’s COVID-19 containment policies have been among the strictest worldwide, including 24-hour curfews that run for two to three days at a time. Borders have been closed to passenger traffic and 14-day quarantine in hotels or purpose-built accommodation has been made compulsory for repatriated citizens.

Jordan has a young population: more than 40 per cent are below 19 years of age. The closure of schools and strict lockdown has interrupted daily life for these adolescents – and interrupted the research we conduct with them.

For Mindset, GAGE’s research partner in Jordan, COVID-19 has required a creative rethinking of the way we collect data. The ethical and public health reasons are obvious: fieldwork requires personal interaction with 100s or 1000s of households. We regularly conduct surveys of vulnerable groups, including Jordan’s expansive refugee population. Should an interviewer or household member be infected, the survey could become a vehicle for the virus.

Still, some of our research is urgent and time-sensitive, like evaluations of eligibility for cash transfers. These projects cannot stop for the pandemic, so we have had to rapidly adapt to our new reality.

The second round of quantitative data collection was scheduled to launch in February 2020. Like the first round, it aimed to collect data on adolescents aged 10-19 and their caretakers in refugee camps, informal tented settlements (ITS), and host communities. However, in-person fieldwork was put on hold due to the pandemic. Mindset proposed a hybrid solution that involved virtual sessions for a sample of respondents and phone interviews for the rest.

To implement remote data collection, we pivoted from in-person interviews to virtual video sessions.  To ensure participation, we needed to offer respondents a system that did not require them to install an external program on their personal devices. In other words, we needed a click-to-play tool. This would accommodate different levels of technology aptitude and would not be limited by operating system compatibility issues.

We adapted a pre-existing online education platform to fit our video call needs. The online platform is an HTML5-based web application that runs using the interviewees’ web or mobile browser, unlike most video conferencing platforms which need plugins or installation to work. Video call support was a key advantage of this application, as video calls fostered a greater sense of comfort and mimicked in-person interviews. Interviewees only had to click on the provided link to be, automatically directed to the interview. The platform also allows live monitoring of interviews and provides each session’s metadata such as time and duration. However, some respondents without smartphones or stable internet connections were unable to use this platform. Others refused to participate because they did not want to be recorded on video.

For respondents who were unable or unwilling to participate in video sessions, we administered surveys over the phone. We set up a remote call centre that allowed operators to work from their homes. The remote call centre recorded each outgoing and incoming call, and provided real-time listening for immediate feedback and mitigation measures for quality assurance.

Both of these strategies allowed us to continue collecting data in difficult circumstances, but they also presented new challenges. In face-to-face interviews, interviewers can be more confident that respondents have privacy during the interview process; this was a particular challenge for the current round of data collection because more household members may be present in the home than usual due to the pandemic restrictions. In-person interaction can also help respondents feel comfortable speaking about sensitive subjects, like sexuality or politics. In this survey, the older cohort of adolescents aged 15-19 were asked questions about sexual violence and marriage. A large number of these surveys were incomplete, as respondents dropped the phone call or refused to continue participating after these questions were asked. No matter what technology we use, digital platforms do not replace the comfort of face-to-face interactions, especially when interviews cover difficult, personal subjects.

Sensitivity isn’t the only problem with administering long questionnaires on the phone: respondent engagement is also difficult to maintain. This was particularly true for the younger cohort of adolescents aged 10-14, who struggled to stay engaged during remote interviews. The surveys were originally designed for a duration of around forty-five minutes to an hour. To accommodate the challenges of conducting long interviews remotely, the surveys were reduced to around thirty to forty-five minutes. But even with these adjustments, the completion rate for face-to-face interviews (99.7%) was higher than the completion rate for virtual interviews (93.6%).

Additionally, because this round of data collection had to be rescheduled, interviews clashed with both Ramadan and end-of-school exams. And while in-person interviews included the incentive of food packages, no incentives were provided during remote interviews. All of these factors made respondents difficult to reach. To address these scheduling issues, two calling teams were established: one team called respondents to obtain consent and schedule interviews, while the other administered the interviews.

As of August 2020, in-person data collection has been allowed to resume by the Government of Jordan. The COVID-19 situation has been largely contained and life has mostly returned to normal. Some partners still choose to collect data remotely, but in projects in which interviewer observation is needed such as vulnerability assessment visits, in-person interviews are still necessary.

In these circumstances, stringent protocols are applied. For example, if the project requires appointments before in-person interviews, we ask the interviewees to wear masks and gloves. Wherever possible, the interview will be completed at the door of the house.

Although COVID-19 has presented many challenges in ethical and accurate data collection, it has also enabled us to build our capacity in virtual research methods. As the situation evolves, we will continue to adapt our practices and collect data as safely as possible.