Replicated Preference Testing to Diagnose Consumer Segmentation
Ennis, D. M. (2017). IFPress, 20(4) 3-4.
In blind consumer preference testing, consumers do not always make consistent choices. This situation will most likely occur when the products tested are similar or exhibit significant variability. The variability may arise from the products themselves, called stimulus variability, or from consumer perception of the products, called neural variability. Consistency in preference responding can inform the degree of segmentation among consumers. Diagnostics for degrees of preference segmentation can be valuable before conducting a large-scale category appraisal. A useful, simple diagnostic for preference segmentation was discussed in a previous technical report on identicality norms. In that report, it was shown that preference segmentation can be diagnosed by comparing preference counts, obtained from a ballot with a no preference option, to expected results if the products are identical. In the same report it was also shown how to establish identicality norms by conducting preference tests on identical products or by predicting the norms using a Thurstonian model of twoalternative choice data. A different way to study potential segmentation is to use replicated testing, which is the subject of this technical report.
Figure 2. Four shapes for the beta distribution for the preference probability among consumers. The black line shows a preponderance of preference for Product A, the green line the opposite, the red line shows that consumer preferences vary symmetrically around 0.5 with few extremes in either direction, and the blue line shows the opposite.