mood was improved following a low-fat, high-carbohydrate breakfast
vs. either a medium- or high-fat breakfast. In both studies “optimal mood
and performance were associated with the meal which was closest in
macronutrient composition to the subjects’ typical intake at that partic-
ular time of day.” Many of these effects occurred within 30 min of
eating. Rogers (1995) suggests that preabsorptive and/or early postab-
sorptive mechanisms are implicated. Currently, the mechanisms at work
here remain unclear, although the hormone cholecystokinin (Stacher et
al., 1979, in Rogers, 1995) is thought to provide part of the mechanism
mediating effects of meal composition.
SENSORY ATTRIBUTES THAT INFLUENCE PREFERENCE
Preference Mapping
Preference mapping offers the potential to link consumer preference
patterns to sensory properties perceived by a trained panel. Preference
mapping is a form of multidimensional scaling in which all consumer
preferences are represented (Carroll, 1972). In a normal preference map-
ping trial, 30 or more consumers are recruited to assess a number of
products for overall preference on a 9-point Hedonic scale. Unlike uni-
variate analysis, preference mapping is a multidimensional algorithm,
and scores for products are not averaged over consumers, but each in-
dividual is represented on the map (Earthy, 1996). A good preference
map that represents a high proportion of variance among consumers may
be used to reconstruct each subject’s approximate order of preference
for the set of products. This is accomplished by orthogonal projection
of the products onto the direction of each consumer vector (Greenhoff
and MacFie, 1994).
Two types of preference mapping can be used: internal or external
(Carroll, 1972). Internal preference mapping is based solely on the pref-
erence scores and provides a multidimensional representation of the stim-
uli. With external preference mapping the preference scores derived from
the consumer panel are projected onto a map that has been derived from
other quantitative information about the products. This information can
be in the form of sensory descriptors, physical or chemical data (Green-
hoff and MacFie, 1994). For both types of mapping the position of each
consumer on the map represents their own direction of increasing pref-
erence; this is known as the Vector mode. This method has several ad-
vantages over traditional methods of analyzing preference data, not only
Sensory Attributes that Influence Preference 161