OPINION10 August 2020

The ‘new normal’ looks almost normal

Covid-19 Opinion UK

Ipsos Mori’s Andrew Green and Neil Farrer reflect on the resumption of face-to-face research and discuss additional measures, alternative modes and interviewers’ perspectives.

Traditional Georgian terraced house with red front door

By definition, face-to-face interviewing involves people getting up close and personal. So, when Covid-19 hit, it was one of the first casualties in the industry. Ipsos suspended its face-to-face operation in line with MRS advice as the lockdown started and furloughed its interviewers.

The questions then became: when could interviewing resume? Would we be able to conduct interviews in the same way as before? If not, what changes would be needed? And how might they affect the data we collected?

We have spent our time during lockdown working to define the optimum conditions that would allow us to return to interviewing. We have liaised with colleagues from other countries who were ahead of the UK on the recovery path.

Early on, we determined that interviews would have to be conducted on the doorstep, not inside homes – a particular challenge for ‘in-home interviewing’. Interviewers must undergo an initial health assessment and regularly confirm they are well and free of any Covid-19 symptoms. CAPI equipment must be cleaned after each interview with any materials provided to participants virus-free. To return, each survey required a formal risk assessment to determine that the interview and any associated tasks could be administered safely.

Alternative completion modes involving online, telephone and video interviews were developed to complement face-to-face. So, a participant uncomfortable with face-to-face contact or who identifies via screener questions asked at the start of the interview as being clinically vulnerable, could still take part in another way. This is important for preserving sample quality and maximising response. These modes will also help interviewers when the weather is inclement so that they don’t need to stay on the doorstep to complete an interview.

We also needed back-up plans in case face-to-face interviewing again becomes difficult or impossible. Localised temporary lockdowns can be managed by more flexible working strategies within areas. However, the possibility of wider restrictions over a longer period has led us to test other modes, with online ‘push-to-web’ proving a viable alternative on some surveys. This can be introduced quickly and targeted to areas where restrictions may arise. With this in place, our research can continue through these restrictions.

Following a successful proof of concept trial in June, we announced that we were ready to return on 3rd July, and we have launched a testing programme to assess these new approaches. We were keen to understand how the new rules would affect things. Could we complete the same number of interviews every day? Would we reach a representative cross-section of the population and collect the same high-quality information?

Early results suggest that interviewing is working well. To some extent, this has been helped by people being at home more than in normal times. We have also found both interviewers and participants adapting quickly to the new procedures.

Refusals have not increased. Those needing to self-isolate on medical grounds were mostly happy to participate in another mode. No issues were encountered with the age profile of participants – even among those over 70 who are most vulnerable. Encouragingly, early results are showing key survey metrics are credible and not too different to those we were achieving before lockdown. Larger field tests are now in progress.

One interviewer told us: “It felt completely normal and everyone was so nice – it felt good to be back in the game. I had two people who were shielding but they both wanted to do it, so I did it over the telephone.” Another said: “I was quite apprehensive before I went out, more so from how participants might not want to do it. But I got exactly the opposite. They're very willing to do it, so it’s not a real problem at all.”

Participants asked about their experience echoed these sentiments. One told us that our interviewer had been better at maintaining social distancing guidelines than the postman or other visitors to her door.

Several told us how much they preferred talking to human beings over other methods. All said they felt quite safe. One participant declined a face-to-face interview and opted for a telephone interview. She was impressed that the interviewer had returned immediately to his car and carried out the interview from there.

Good participation rates and high-quality insights are particularly important for audience measurement studies we run on BARB, RAJAR, PAMCo and Route. All use face-to-face recruitment and data collection due to the need for robust, credible, accurate and (as far as possible) beyond reproach figures to inform the buying and selling of advertising.

Face-to-face recruitment and data collection represents the gold standard. It enables the highest quality probability sampling base, using household addresses as its foundation. Households are a far more solid foundation than, say, lists of telephone numbers or email addresses. In the interview itself, the presence of an interviewer helps prompt more considered responses to questions, better engagement of participants and minimal confusion over the meaning of questions. Longer and more complex surveys can be conducted.

End users of face-to-face generated data are right to want to understand any potential impact this new approach will have on their results. So far, it looks as if the ‘new normal’ will feel mostly normal for face-to-face interviewing.

Andrew Green is global business development lead, audience measurement, and Neil Farrer is UK head of audience measurement at Ipsos Mori