Temporal liking and CATA analysis of TDS data on flavored fresh cheese

- Analysis of temporal liking data relative to temporal dominance of sensations (TDS) on the same products.
- Consideration of TDS data (per time interval) as check-all-that-apply (CATA) data.
- New analysis of TDS data using methods primarily proposed for CATA.
- Preference mapping based on TDS (in CATA-format) data.

This paper describes some data analyses for the TDS data used in the workshop at the 2014 Sensometrics meeting, employing flavored fresh cheese. An initial statistical significance test for the data of both the panel and the consumers shows that the consumer data is much richer in terms of discriminating between the products. This might be attributed to the lower sample size for the trained panel, but even increasing the significance level to 20% (rather than 5% as for the consumer data) did not give similarly rich information. The panel primarily discriminated the products on a single attribute. Adding the temporal hedonic data to the difference TDS curves between products suggests that absence of cooked herbs and the presence of fresh herbs seem to be main drivers of liking. For an alternative approach, the data was split into 4 time periods, and considered as check-all-that-apply (CATA) per period. The resulting Penalty-lift analysis supports the above findings and indicates that garlic and salty are the next most important positive, with pepper being another negative driver of liking. The respective correspondence analysis (CA) shows a clear separation between the products. Results suggest that it might be worth exploring repeated CATA as an alternative to TDS, offering the possibility to check more than just one attribute at a time. Finally, consumer (temporal) liking was added to the CA in a biplot way, showing substantial heterogeneity among the assessors.