Archive for October 20th, 2009

Consistency of online samples is a core issue for market researchers. If your data changes, it is essential to know if the changes are real or the inadvertent product of sample inconsistency.  One, now historic, example of this happening was presented by Ron Gailey, now of Coca Cola, but previously of Washington Mutual, who disclosed how 29 studies representing 40,000 online interviews had gone astray due to panel inconsistency. In the WaMu research, the change was due to shifts in respondent tenure that resulted from changes in the panel’s constituents over the two-year span of the base research. Gailey’s research showed a 30% drop in buyer demand for WaMu’s financial products; a result (2006-7) unsupported by sales. His conclusion, after much study, was long-term panel members were less optimistic about their purchases than new panel members. Others have since corroborated these findings. The lingering question, now that WaMu is gone, is how the tainted research impacted on critical business decisions.

You can read the full article published in MRA’s Alert! Magazine October 2009: Consistency: the New Quality Concern, by Steve Gittelman and Elaine Trimarchi