Differences in citation proportions in CATA questions can be interpreted as differences perceived intensity of sensory attributes

Abstract

CATA (check-all-that-apply) questions enjoy high popularity for sensory product characterization with consumers. Amidst extensive methodological characterization, the meaning of term citation frequency has received modest attention only, possibly hindering further uptake. The present research performed a more detailed exploration of the relationship between CATA responses and perceived intensity of sensory attributes. In addition, the following question was addressed: How many consumers are necessary to interpret differences in CATA citation proportions as a proxy of differences in perceived intensity? The present research includes new empirical data from four consumer studies conducted in New Zealand and China under central location test conditions and a re-analysis of one consumer study. When dealing with samples with clear differences, the percentage of consumers selecting a sensory term for describing samples tended to be linearly correlated to average intensity ratings. A median of 15 consumers was needed for CATA to show good linear correlations with average intensity ratings, and the minimum number of consumers needed to reach the threshold for a good correlation was generally lower than 30. Practical ApplicationsResults add to existing guidelines and say that for sets of sample sets with at least moderate differences in a specific attribute, a minimum of 50 consumers is necessary to obtain reliable sensory characterizations using CATA questions and interpret differences in citation proportions as a proxy for differences in the perceived intensity of that attribute. However, practitioners should be aware of the limitations of CATA questions for measuring attribute intensity, such as studies involving complex terms or samples with subtle intensity differences.

How do CATA questions work? Relationship between likelihood of selecting a term and perceived attribute intensity

“…As perceived terms intensity increased, likelihood of selecting a CATA term tended to linearly increase. This matched results from previous research by, for example, Bruzzone et al (2012), Reinbach et al (2014), Oliveira et al (2018), Jaeger, Chheang, et al (2020), Choi and Lee (2019), and Vidal et al (2021).…”

“…This suggested that CATA questions may be unlikely to discriminate among samples with similar and high intensity of a given term/attribute. In such situations, the uses of RATA (rate‐all‐that‐apply) questions may be necessary, as recommended by Vidal et al (2018, 2021). Briefly, in RATA questions, participants select if a given term applies or not to describe a given sample, and if it does, they rate the perceived intensity on an intensity scale.…”

“…Following the recommendation of Vidal et al (2021), the present research extended beyond the use of CATA questions for sensory product characterization. This was warranted because CATA questions are also being used to characterize other aspects of eating and drinking experiences notably emotional associations, conceptualizations, and situational appropriateness (e.g., Cardello et al, 2022; Jaeger & Ares, 2022).…”

“…Despite their simplicity, CATA questions have been reported to accurately discriminate among samples (Jaeger & Ares, 2022). Several studies have reported that citation frequency is a proxy for perceived intensity, albeit not a direct measure hereof (Bruzzone et al, 2012; Choi & Lee, 2019; Jaeger, Chheang, et al, 2020; Oliveira et al, 2018; Reinbach et al, 2014; Vidal et al, 2021). If for a given term (X), the citation frequency for two samples (A and B) are X A and X B , then the average intensity of X can be inferred to be higher in A than B if X A > X B .…”

Identifying sensory drivers of liking for plant‐based milk coffees: Implications for product development and application

“…On the other hand, the major sensory perceptions of the dairy milk coffee sample (P12) were milky, smooth, thick, and bitter. Vidal et al (2021) analyzed the empirical data from five consumer studies (each with 159-209 participants) and concluded that the differences in citation proportions in CATA questionnaires could be used as a proxy for the differences in the perceived intensity of sensory attributes under certain criteria (i.e., moderate differences among samples, ≥50 panelists, and no involvement of complex terms). There was a significant difference in the selection frequency for milky flavor between plant-and dairy-based milk coffees (Table 1).…”