A Comparative Study of Open-Comments and CATA to Identify Drivers of Liking and Disliking of Milk Chocolates

Abstract

Collecting Open Comments (OC) is a common way to learn about the product features that consumers like and dislike in consumer research methodologies such as preference mapping. OC has the advantage of being unbiased and providing a causal answer to the question "What do you like/dislike about this product?". However, encoding the information they provide is a time and resource consuming mandatory pre-processing step, whether performed manually or with a dedicated AI algorithm. In this context, the Check-All-That-Apply (CATA) method can be a valid and less expensive alternative.

We conducted a consumer study on 8 milk chocolates with 361 respondents to compare the information (liking scores, liking drivers, and liking clusters) obtained from OC and CATA. Each consumer tested 6 milk chocolates out of the 8 milk chocolates based on a MOLS design and provided for each chocolate a liking score (7-point scale) and reasons for liking and disliking using either OC (n=204) or CATA (n=157) with 28 terms.

We found no notable differences between the liking spaces obtained from the CATA and OC groups, and the same two liking clusters (mild and strong chocolate lovers) were identified in both groups. The drivers of liking and disliking associated with each product were largely consistent in both groups. However, some subtle differences in citation rates and attitudes were noted. In OC, consumers tend to disregard appearance and texture more in favor of taste than in CATA. In addition, the drivers of disliking provided by OC were less diverse and less frequently cited than in CATA. Conversely, OC showed specific shape-related terms that were not present in CATA.

This suggests that CATA is a promising alternative to OC, as the overall picture was the same, but there are still some slight differences between the methods.