SELECTION OF FRAGRANCE FOR COSMETIC CREAM CONTAINING OLIVE OIL 163 in shopping areas, universities, and other public places. Respondents were asked to select up to 3 of 28 fragrance names that they considered most appropriate for each of three cream types. The list of fragrance names covered a wide range of notes generally associ- ated with cosmetic creams, such as fl oral, fruity, citric, herbal, and spicy notes. The cream types were hand cream, body cream, and facial cream. The fragrance names were odorless, apple, azalea, blueberry, cherry, coconut, fresh fruit, honey and lemon, jasmine, kiwi, lemon, linden and magnolia, magnolia, melon, olive and sage, orange, orchid, peach, pear, pineapple, pink grapefruit, rose, rose and lemon verbena, strawberry and blackberry, tropical fruit, vanilla, white fl owers, and wild strawberry. Frequency of mention was determined for each note and the six notes mentioned most frequently overall were identifi ed. On the basis of these six fragrances, essences that would contain the notes mentioned were screened. Lariales S.A., a local company, kindly pro- vided the essences used in this study, as well as advice on their selection. Six essences mentioned in Table I were selected, as containing the most representative notes named by the consumers. On the basis of these fragrance names selected by consumers in the preliminary study, essences were compounded and incorporated into a cream base, and were then subjected to an olfactory test using women consumers. OLFACTORY TEST WITH WOMEN CONSUMERS Samples. A cream base was prepared with acrylate–acrylamide copolymer, Picual extra virgin olive oil, propylene glycol, methyl and propyl paraben, aqua and butylated hy- droxytoluene. The fruitiness intensity of the olive oil used was below 3.0 on the IOC scale (IOC/T.20/Doc. no 15/Rev. 4, 2011) (21), according to Gámbaro et al. (22). The emulsion was prepared by mixing all the ingredients followed by 5-min stirring (Servodyne Mixer Head Model No. 50003-45, Cole–Palmer Instrument Co., Vernon Hills, IL) at 500 rpm. Fragrant essences were added to this emulsion to a fi nal concentration ranging from 0.4% to 1%, following the supplier’s instructions. The six cream samples were labeled as CE1– CE6, where C indicated the cream type used. Table I Essences Selected for This Study Code Principal note Other notes E1 Vanilla Lime, sweet orange, bitter orange, cocoa, caramel, coconut, and ambergris E2 Lemon Lime, bergamot, orange, jasmine, violet, pineapple, musk, and vanilla E3 Fresh fruits Orange, bergamot, peach, green notes, musk, and vanilla E4 Jasmine Violet, white fl owers, vanilla, and sandalwood E5 Rose Silver wattle, violet, vetiver, sandalwood, ambergris, and musk E6 Linden/magnolia Herbs, gardenia, musk, and cedar

JOURNAL OF COSMETIC SCIENCE 164 Consumer test. A total of 63 female consumers of cosmetic creams aged between 21 and 72 were recruited randomly in shopping areas, university facilities, and other public places in Montevideo, Uruguay. Taking into account the areas where the respondents were re- cruited, the sample was assumed to represent the general Uruguayan middle-income groups. The test was conducted at the site of recruitment, exercising care to avoid odor contamination and to create the conditions for respondents to perform the test in a re- laxed atmosphere. Respondents were presented with 20 g each of the six cream samples in white plastic cups coded with three-digit random numbers. The sequential monadic sample pre- sentation protocol was used, using a different random presentation order for each respondent. The respondents were asked to remove the cup lids, smell each cream sample, and rate it according to overall liking on a 9-point structured hedonic scale ranging from extreme dislike to extreme liking, and intention to purchase on a 9-point structured hedonic scale ranging from defi nite unwillingness to defi nite intention to purchase. No information about the name of the sample fragrance was provided to the respondents. Finally, the respondents answered a CATA question consisting of a list of 32 terms from which they selected those applicable to each of the six samples they had smelled. The CATA terms could be classifi ed into the following categories: Odor: delicious, disagreeable, strong, mild Affective: must-have, glamorous, for pampering oneself, energizing Effect of cream on the skin: nourishing, moisturizing, softening, beautifying, anti-aging, anti-wrinkle Price: cheap, expensive Target market: young women, older women, exclusive, mass market Zones of application: feet, hands, face, body Occasions of use: summer, winter, day, night Other: fresh, healthy, natural, artifi cial Respondents also completed a brief survey of sociodemographic data (age, marital status, number of persons in the household, number of children in the household, and highest educational level attained). They were also asked about their consumption frequency of face creams (moisturizing, nourishing, and anti-aging) and body creams (moisturizing, nourishing, anti-aging, slimming, and toning creams) with structured answers (three op- tions): never, sometimes, or always used. Data analysis. An analysis of variance (ANOVA) was performed on the overall liking and intention to purchase data using sample as the variation factor. Signifi cant differences between means were determined according to the Tukey test (p ≤ 0.05). CATA question. For the CATA question, frequency of mention by respondents was counted for each attribute and sample. To detect differences in consumers’ perception of the eval- uated fragrances, Cochran’s Q test was carried out for each of the 32 terms considering sample and consumer as variation factors. Cochran’s Q test is a non-parametric statistical test used in the analysis of two-way randomized block designs to determine whether k treatments have identical effects when the response variable is binary.