A simple process to analyze free sorting data given incomplete block designs
The free sorting task, which involves asking each subject to group together stimuli based on their perceived similarities, is widely used in the sensory world. However, industries need to evaluate more and more products and this task shows its limits when the number of products increases, since it c...
Saved in:
| Published in | Science Talks (Online) Vol. 10; p. 100336 |
|---|---|
| Main Authors | , |
| Format | Journal Article |
| Language | English |
| Published |
Elsevier
01.06.2024
|
| Subjects | |
| Online Access | Get full text |
Cover
Loading…
| Summary: | The free sorting task, which involves asking each subject to group together stimuli based on their perceived similarities, is widely used in the sensory world. However, industries need to evaluate more and more products and this task shows its limits when the number of products increases, since it can cause fatigue to the subjects. To address such a limitation, one recommendation is to resort to an incomplete free sorting task. However, the majority of methods for analyzing free sorting tasks that require complete data can no longer be used.Among the existing methods to analyze free sorting data, Cariou & Qannari (2018) proposed setting up a product x product contingency table and then performing a Correspondence Analysis on this table to determine a product map. We propose to extend this approach to the case of an incomplete free sorting plan by simply reporting in the contingency table zero occurrences for the products that have not been with the others. This straightforward strategy is suitable for incomplete block design by following three instructions.This adaptation is illustrated on a case study of wine aromas where subjects performed complete and incomplete free sorting tasks. Finally, a comparison of the results is made. |
|---|---|
| ISSN: | 2772-5693 2772-5693 |
| DOI: | 10.1016/j.sctalk.2024.100336 |