.50 b 4 "" i{ 46 H 44 G 4 :• L L .'3_,2 .30 ""P I f.: I HE':' S t.l.: L I K I Iq G (TOTRL PAIqEL) OUERALL Ii I ii 60 7 0 ß --, I ½: I H F.. o E R A L L 60 SF'I •:: I NESS ( S E N ':' I'1R "r' L I K I NG OUERALL SEGMENTS ) 55 OL4 I GH r,IED 50 45 40 •5 .'3_0 ,.0 . 30 40 50a 60 " ':'P I C: I N E""-' Figure 6. Figure 6A--Fitted curve for liking vs. spiciness, for the total panel. Figure 6B--Fitted curves for liking vs. spiciness, for the three sensory segments.

SENSORY SEGMENTATION OF FRAGRANCES 245 Table VI Demographics of the Segments (% of Total) A. Education Total Low Meal High Grade School -- Some High School 6 7 5 5 Graduated High School 31 35 33 18 Some College 36 35 35 39 Graduated College 17 11 21 21 Post Graduate 11 13 6 16 B. Income Total Low Meal High Under $10,000 6 6 6 8 $10,000-$15,000 11 8 17 5 $15,000-$20,000 11 11 13 8 $20,000-$25,000 11 17 6 8 $25,000-$30,000 17 15 21 16 $30,000- $35,000 14 14 13 18 $35,000-$40,000 6 10 3 13 Over $40,000 23 19 21 24 C. Age Total Low Med High 18-34 47 47 54 36 35 -49 53 53 46 64 D, Employment Total Low Meal High Full Time 32 33 33 26 Part Time 25 25 16 39 Not Employed 43 42 51 34 E. Frequency of Using Fragrance Total Low Med High Daily 54 58 49 55 5-6 Times a Week 24 19 30 24 3-4 Times a Week 22 21 21 21 F. Type of Fragrance Used Total Low Med High Spray On 43 51 41 32 Splash On 6 7 6 5 Both 50 42 52 63 Note: Percentages may not add to 100% because of rounding. 4. These segments differ in the way that sensory attribute levels drive overall fragrance liking, for "blind" fragrance evaluation. 5. There are some demographic differences (especially age) among the segments. DISCUSSION DO SENSORY PREFERENCE SEGMENTS REALLY EXIST? The thrust of this paper is to uncover the existence of sensory segments in the popula- tion. Do they really exist, however, or are they statistical artifacts?