COVID-19 Impact: Will our panels change?

By: Pete Cape
Director, Global Knowledge

I wrote in my first blog on March 26 on research in the time of COVID-19 why I believed our samples wouldn’t fundamentally change.

Having had time to look at some hard data, I can tell you we are not observing large upward changes in what we call “starts,” that is panellists turning up in our systems looking to do a survey. In fact, looking at the Year-on-Year data for Spain, currently in the grip of the pandemic, the data is almost identical now to the same period in 2019, and to 2018. Mostly of course this is because we actively manage our invitation rate to match our demand. So, if there is an increased demand to take surveys then we should see it in an increased conversion/click-thru’s on our invites, and we don’t.

An associated question has started to rear its head – are we seeing increased join rates as people are laid off from their jobs or just have more time on their hands?

My gut response to this question was, well “gut”:

If anyone thinks they are going to make up for a lost salary by joining an online panel and doing loads of surveys they are sorely mistaken. They can make more money stocking shelves in their local supermarket, pulling beans or picking fruit on the farm or basically anything but doing surveys.”

But I thought I should check with the teams responsible for the panels in case my gut was wrong. It wasn’t. Recruitment and refresh rates are the same now as they were this time last year.

What is undeniable as we go through the pandemic is that incomes of some panellists may well be hit, and the impact of it dependent on the actions taken by national governments to support incomes. This will potentially have a “knock-on” effect on data.

It is more important than ever to ask carefully thought out questions about income depending on the survey subject. If the survey relates to current or near future expenditure, then an accurate estimate of current actual income is required. If the survey is about a far future or perhaps hypothetical purchase then a measure of “normal” income might be much more appropriate, along with a sensitive note about coronavirus. To decide what to do you need to think about why you are asking this question – what does it tell you about the people and/or their household about their lives and aspirations? And note that balancing to “nat rep” may become problematical as the national picture changes faster than census bureaux can keep up and produce statistics on. This is a time to be pragmatic not dogmatic in survey planning, execution and analysis.

Click here to download Dynata’s Global Trends Report Special Edition: COVID-19