Incidence is a phrase that comes up a lot in research but it isn’t often understood immediately outside the research industry. So let’s have a look at incidence and how it relates to online panel research.
What is incidence?
In a nutshell, incidence is the percentage of respondents from a total universe who meet various requirements. In the world of research, incidence is usually defined as the percentage of people who meet the survey sample requirements. In terms of online panel research, incidence is the percentage of people in an online panel that meet your sample requirements and are therefore eligible to complete your survey.
Your requirements affect the incidence
For example, if your online panel research project calls for UK females aged 18-45 and out of 1000 invited, there are 100 people who match those requirements, the incidence is 10%. Most panel providers already know these basic demographics about their panellists, however, so they are able to deploy surveys only to those who meet the specific requirement. The incidence in this case therefore goes back up to 100%.
When your respondent profile requirements are more complex, however, it can be difficult to know in advance what the incidence will be. The incidence has a direct bearing on the cost per response.
Online panel research projects with less complex sample requirements will generally yield a higher incidence. The higher the incidence, the lower the cost per response and the easier it is to determine an accurate budget for your online panel research project.
Planning your online panel research depends on the incidence
When dealing with complex sample requirements, it can be harder to predict the potential incidence. In these cases, it may be an idea to deploy to the panel in phases so you can:
- test the incidence with a small initial sample
- pause the survey
- calculate the incidence at this stage
- have the panel provider calculate the projected cost per response for the remainder of the sample.
This way, you’ll have a much more realistic idea of what the project is going to cost.
Let’s consider, for example, how this would work if you wanted to conduct a survey of UK females aged 25 – 55 who drive a white car and you need 1000 responses in total. Because of the way panellists are profiled, the panel provider may already know which panel members are female aged between 25 and 55 and so will deploy the survey to these panel members. What isn’t known, however, is how many of these panellists drive a white car. By deploying your questionnaire to just 100 of these panellists and using a screening question to allow only those who drive a white car to respond, you discover that 65 of these 100 panellists meet this requirement. This means that the overall incidence for your sample is 65%. The panel provider can then quote on the cost per response knowing that the incidence is going to be around 65% and this makes it much clearer in advance how much your overall sample of 1000 is going to cost.
Incidence is a vital consideration in any online panel research project. Understanding your requirements, and by extension the potential incidence, will help in planning any panel research project. We always encourage our clients to define their sample requirements clearly and discuss them with us when they’re considering online panel research. We can work together to find ways to balance sample size and cost for all research projects.