A New View of Panel Quality
My colleague, Peter Mackey, took part in IIR's Research Industry Summit and was particulary taken back by a presentation on panel quality by our client at Washington Mutual, Ron Gailey.
Ron and his team began studying the impact of panel quality on research results to see how big an issue existed and what the real drivers of inconsistent data were. Three big findings:
1) The longer a person has been part of a panel and the overall number of surveys they have taken has a significant impact on ratings.
2) New panelists and panels tend to give higher scores than long standing panels and panelists
3) The lower scores tend to be more accurate because new survey takers suffer from a "halo" effect
So, what next? Ron states that its essential to understand the makeup of online panels and to adjust to create consistent, balanced sample plans. Furthermore, using a single sample source could be unintentionally biasing your data.
Looks like professional respondents might not be the problem after all!
Comments