There is currently much debate about the role of added sugar in our diet in relation to weight gain and obesity. In particular, sugar-sweetened beverages are thought to be articularly problematic, with some claims that ' liquid calories' are not sensed by the petite control system. Research findings spanning over 3 decades have shown that this not the case - these studies have used a procedure in which participants consume a drink containing sugar, or a control drink (water, or a drink sweetened with a low-energy sweetener), and subsequently eat a meal in which their food intake is measured covelily. This 'preload, test-meal procedure' has shown that consuming sugar in a drink does suppress appetite, but not sufficiently to fully 'compensate' for the calories in that drink. This thesis focused on aspects of study design in relation to the preload, test-meal procedure. Five experiments were conducted (testing a total of349 participants, eating a total of 698 test meals). The energy compensation (EC) for sugar in the traditional (commonly-used) cross-over design was modest when a sugar-sweetened beverage was matched to a low-energy sweetened beverage in terms of flavour, appearance and volume (37%, p = .048). Analyses suggested that this design may underestimate compensation due to a 'carry-over' effect, possibly arising from learning about the satiety effect of the sweet drink served on the first occasion carrying over to influence satiety experienced for the (higher or lower energy) sweet drink served on the second occasion. Consistent with this, compensation was greater when in another experiment water (familiar as nonsatiating, and different in taste and appearance to the sugar-containing drink) was used as a control- EC was 91 % (p = .009). The improved EC in this experiment was more apparent in men (123%) than in women (39%). In two further experiments, the satiety effects of sugar were examined in betweensubjects designs after controlling for baseline meal intake. Using this approach to avoid carry-over effects, the EC of sugar versus low-energy sweetener was 63% (p = .01). Again, EC was greater in men (109%) than in women (16%). This gender difference was not explained by differences in cognitive control of eating between men and women (i.e., women were not more likely than men to endorse restraint, meal planning, etc. as reasons for ending their meal). Another between-subjects design study varying the viscosity of a sugar versus low-energy sweetener preload showed that, contrary to current claims, EC for sugar in a drink was not less than EC for sugar consumed in semi-solid Gelly) and solid (candy) foods. Taken together, this thesis shows that the satiating effect of sugar in a drink has been underestimated by previous research. Nonetheless, it is likely that use of low-energy sweeteners will reduce energy intake, and therefore the risk of overweight and obesity, because they, like the use of fat replacers, reduce dietary energy density.

Emotion research in sensory and consumer science has gathered significant momentum over recent years and the development of effective emotion measurement methods is a priority in this rapidly growing area. The aim of this research was to advance the use of consumer-led emotion lexicons by using focus groups to increase the efficiency of lexicon generation and by decreasing the number of consumer response categories. In parallel, the ability of the newly generated reduced lexicon to discriminate emotional response across different gender and age groups, and across sensorially distinct beer samples, was evaluated. The new approach was largely effective at discriminating across samples and revealed significant differences in emotional response between genders and between age groups. The reduced lexicon was compared to a full lexicon to ascertain their relative efficacies. Whilst there were differences between the two form lengths, neither was convincingly more effective at sample discrimination than the other, although the full form better differentiated between age groups. The reduced form was also applied to cross-cultural comparisons through the generation of a reduced product-specific consumer-led emotion lexicon in Spain. As in the UK, the approach discriminated well between samples and was able to differentiate between consumer groups. Comparing Spanish and UK responses, ratings of emotions associated with pleasure/pleasantness were similar but there were differences in the use of emotions associated with arousal/engagement/activation. This new methodology was therefore demonstrated to be a valuable tool for investigating cross-cultural emotional response. The approach developed in this thesis provides researchers with an enhanced consumer-led emotion methodology for use with food and beverages. As well as being relatively quick, the approach has been proven to differentiate between products and reveal differences concerning emotional response across different consumer groups and between cultures. These attributes make this emotional measurement approach extremely valuable to this young research area.