Survey response rates remain strong during COVID-19
Despite researchers’ initial fears that COVID-19 would impact response rates, the latest data shows research samples remain strong, with respondents clearly willing to take part in both market and academic research during the pandemic.
COVID-19 has undoubtedly caused concerns for researchers over the past 3 months — first, there were fears that the public wouldn’t want to take part in research, plus many had to look to new sampling techniques after methods like face-to-face polling became impossible.
Indeed, Pew has suspended its global polling as face-to-face interviews are a high-risk activity that could increase coronavirus transmission rates.
As a result, many researchers have switched their focus to alternative sampling methods like web panels - they offer plenty of opportunities including probability-based, representative samples in many countries as well as opt-in methods that use quotas or data weighting to produce samples that look a lot like your target population.
Response rates remain stable
While the methods some researchers are using have changed, the public’s appetite to respond has not. Our own research found approximately 86% of people were happy to take part in research during the pandemic, and new data shows that response rates are in fact on the rise.
With many people out of work and staying at home, researchers have found that survey participation is up across modes while online panel providers have reported above normal daily completion rates and a decrease in fraudulent response worldwide.
We recently spoke to several online panel organizations about global sample quality and participation rates during COVID-19. Here’s what we learned:
- Participation initially declined but quickly rebounded in countries that have seen viral hotspots.
- Panel members had fewer opportunities to respond to surveys because research requests declined while organizations paused research. These requests have now begun to rebound.
- Demographically, samples look roughly the same, although worldwide, most surveys are reporting higher levels of unemployed respondents. This corresponds to an increase in global unemployment rates.
- New members are joining online panels at the same rates.
- Even surveys of healthcare providers are showing typical participation rates.
While the new normal of working from home has made general population samples easier to reach, surveys of establishments that require contacting people at work may be more challenging.
Observations for surveys of business and professional respondents have been mixed. We noticed early on in the COVID-19 crisis that it took longer than usual to obtain complete samples for some business surveys. In recent weeks (11+ weeks into remote working for most North American and European countries) things are nearly back to pre-COVID-19 levels. One provider noted that business survey respondents often complete their surveys off-hours, and that no difference in participation rates had been observed in their panel.
Now is the time to use online panels
Given the data on response rates, it’s clearly a great time to conduct research with online panels, particularly as you’ll likely find general population respondents much more quickly than in other times.
Essential workers and business/organizational representatives may be more difficult to find, but they are not impossible — that said, you should consider building extra days into your schedule to allow for these respondents.
To ensure that your study achieves the sample characteristics you want, continue to follow these best practices:
- Use quotas to match key demographic variables such as race/ethnicity, income, sex, age, or region - or weighting after-the-fact - but avoid using both weighting and quotas.
- As appropriate, target your demographic quotas to national statistical estimates such as the U.S. Census Bureau, the U.K. Office for National Statistics, or the German National Statistical Office.
- Avoid or allow extra time in your project schedule for surveys with highly stressed or busy populations in the era of COVID-19: essential services providers, health care providers, parents of children home from school and daycare.
- Conduct an early data review to ensure that your sample coming in is balanced as expected.
Advocating for your academic research during a pandemic
December 1, 2020