Abstract:

The war for online quality is being fought at respondent engagement. So far, the solution appears to be identifying offenders and eliminating them. However, not all satisficers are equal; we need rules of engagement that guide us toward eliminating those respondents who provide us with poorly considered data and have an impact on the deductions and decisions we might make from that data. It is not enough to eliminate the worst offender; some measure of how much damage they cause needs to be included. This presentation addresses the construction of a new metric that includes considerations of satisficing, data shift, and sample size. in a real war there are rules of engagement that dictate under what circumstances commanders can order their troops to fire, here such guidelines are provided for our industry.