Ensuring Reliability

Research conclusions are rarely black-and-white. Conclusions must account for the probability of a “false positive”. That is, the research concludes that there is a difference between groups or a positive effect of a stimulus, but in actuality there is not an effect.

Typically, the standard for research reliability is a false positive rate of 5% (i.e., a 95% confidence level). However, recent research-on-research shows that the actual rate of false positives likely is closer to 45%—9 times higher than the false positive rate is typically believed to be.

Marketing activities based on false positive conclusions may fall short of the business objectives. This may force the business to increase the marketing investment and extend the marketing activity schedule in order to reach the objectives. The increased investment may drive down marketing ROI, and the extended activity schedule may allow competitors to encroach on the market opportunity, degrading the potential return.

Small increases in reliability may provide a large payback by increasing marketing ROI and avoiding schedule extensions.