SELECTION OF FRAGRANCE FOR COSMETIC CREAM CONTAINING OLIVE OIL 171 As mentioned above, a correspondence analysis was used to visualize the relationships between products and associations. The resulting perceptual maps are shown for the two clusters in Figures 1 and 2, respectively. The fi rst two dimensions of the correspondence analysis explained 61.2% and 79.2% of the variability of the experimental data for Clus- ters 1 and 2, respectively. Cluster 1 respondents differentiated between three groups of samples (Figure 1). On the right are samples CE3 and CE2, with fruit and lemon fragrances, characterized in terms of “summer use,” “for young women,” “daytime use,” “fresh,” and “energizing.” Below is sample CE4, a complex fragrance with several fl oral constituents, including jasmine and white fl owers, as well as vanilla and sandalwood. CE4 is characterized in terms of “facial use,” “expensive,” “nourishing,” “softening,” and “glamorous,” consis- tent with its high overall liking score. On the left are samples CE1, CE5, and CE6, with fl oral and vanilla fragrances, characterized in terms of “cheap,” “use on hands,” and “night time use.” Figure 2 shows that Cluster 2 respondents differentiated between two groups of samples, with a lesser degree of discrimination than Cluster 1. On the right of the fi gure are samples CE2, CE3, and CE5, characterized in terms of “summer use,” “for older women,” with “disagree- able,” and “strong” smells, “artifi cial,” “cheap,” “for a mass market,” and “use on feet.” On the Figure 1. Correspondence analysis plot for CATA terms associated with the different types of fragrance by Cluster 1 respondents.

JOURNAL OF COSMETIC SCIENCE 172 left are samples CE1, CE4, and CE6, characterized in terms of “nourishing,” “softening,” “fresh,” “moisturizing,” “delicious smell,” “natural,” “use on body,” and “night time use.” Fragrance E4 was selected for the further development of a cosmetic cream as it had the highest overall liking scores among respondents of the two clusters and was signifi cantly associated with “nourishing,” “moisturizing,” “softening,” “delicious smell,” and “mild smell,” as well as considered a suitable fragrance for face and body creams, and associated with a “natural” image. CONCLUSIONS Use of “CATA” questions allowed the identifi cation of different perceptions of six fra- grances among consumers. The results provided an insight into consumers’ perceptions of different aspects of creams associated with their fragrances, including their possible effects on the skin, affective associations, zone of application, and target market. The selection of a suitable fragrance can contribute to reinforcing product image, and the use of CATA questions enables the rapid identifi cation of associations made by consumers with a cream’s fragrance, as well as providing useful insight on possible marketing and communication strategies. ACKNOWLEDGMENT We thank Lariales S.A. for advice on selecting the fragrant essences used in this study. Figure 2. Correspondence analysis plot for CATA terms associated with the different types of fragrance by Cluster 2 respondents.