Since we get quite a few questions about incidence, we thought we’d put together an article addressing it specifically. Feel free to add your comments!
When talking about online sample, incidence (or incident rate) is always part of the discussion. Dictionary.com defines incidence as the rate or range of occurrence or influence of something. When we talk about incidence we are referring to how hard it is to reach a specific audience. The survey method impacts incidence, e.g. it is easier to reach internet users online versus telephone. Another simple explanation of incidence would be the number of people who would qualify out of 100. For example if you are looking for those who drive cars and 80 out of 100 people drive a car, your incidence would be 80%. The equation would read like this – Incidence = # of people who qualify / (# of people who qualify + # of people who do not qualify). If you have already conducted a study and want the incidence, the equation would be Incidence = # of people who completed / (# of people who completed + # of people who screened-out). 100 completes + 150 screen-outs or terminates would equal a 40% incidence (100/250=40%).
Many times the impact of qualifiers on your incidence can be confusing. First, let’s talk about the term targetable. Targetable means profile information that is pre-identified such as age, gender, income and more depending on the company that you work with. For instance in a study looking for adults age 18+ who have migraines, the qualifiers are age and migraine history. If migraine history is profiled within the panel then that is a targetable attribute. In this case the incidence would be 100%. Another example would be a study looking for adults age 25-54 who read the NY Times. We can target age and newspaper readers, but not necessarily what newspaper they read. An estimate of what percentage of newspaper readers specifically read the NY Times would be the incidence, e.g. if we assumed 10% of newspaper readers read the NY Times, the incidence would be 10%.
Sometimes you will hear the term Net Incidence. This comes into play when a study has multiple quotas on Non-targetable information. Net Incidence refers to the overall incidence including quota groups and segments. An example of this would be in the newspaper reader study having quotas on the number of completes for NY Times, Wall Street Journal, and USA Today readers. When one of the quotas close, the overall incidence decreases for the study. At the end of the study a net incidence would be figured by # of completes / (# of people who completed + # of people who screened-out including quota full terms).