Bias | Know the Pitfalls

There are three types of biases that should be avoided at all costs: the halo effect, first-sample bias, and adaptation. To avoid the halo effect, it’s best to have panelists complete their evaluations individually before opening things up to discussion. It’s helpful to foster a culture in which all opinions are welcome and valid. Calibration samples and randomization are good ways to help nullify the first-sample effect. To avoid adaptation bias, it’s helpful to keep sample sizes relatively low—perhaps eight samples in a sensory session. Keep it focused.