the plate and food reduced over- and underserving bias on both small and large
plates (Van Ittersum and Wansink 2012 ).
In addition to size or perceived size of the dishware, the shape of a container can
give deceptive cues as to the volume of its contents. People overestimate the
volume of the contents in a tall cylindrical glass but underestimate quantity in a
short, wide glass. Experimental evidence indicates that, on average, people will
consume 25–30% more from a short, wide glass and, again, not be aware of the
increased intake (Wansink 2010 ). Both college students trained to pour shots and
professional bartenders with an average of six years of experience poured more into
short, wide glasses: 59.6 and 49.7 ml compared to tall, narrow glasses of the same
volume: 44.9 and 45.4 ml (Wansink and van Ittersum 2005 ). Another optical
illusion may account for the overestimation of the quantity of liquid in a tall glass.
The vertical–horizontal illusion shown in Fig.10.2is the tendency for observers to
overestimate the length of a vertical line relative to a horizontal line of the same
length (Robinson 1998 ). This happens even when people are aware that the lines are
of the same length. Wansink reported that people routinely estimate that a vertical
line is 18–20% longer than the same length horizontal line (Wansink 2010 : 61).
Susceptibility to the vertical–horizontal illusion occurs more frequently among
people from‘Western’cultures with bottles, glasses, and cartons as individual food
serving containers compared to those living in‘eastern’cultures with cups, bowls,
and boxes (Shiraev and Levy 2007 ).
Misperceptions in estimating portion size is only one aspect that contributes to
difficulties in obtaining accurate dietary data. There are manufactured deceptions of
food processing (i.e., foods sweetened with non-caloric sweeteners, the mouthfeel
of fat in fat-free foods), packaging, and dinnerware. Furthermore, people simply do
not remember what they ate even when the window of recall is only a few minutes
or hours. This forgetfulness or informant inaccuracy is a major drawback of tra-
ditional retrospective dietary studies using either food frequency questionnaires
(FFQ) (e.g., Havard/Willett FFQ, National Cancer Institute/Block FFQ) or the‘gold
standard’24-h dietary recall (Bernard 2011 ; Bernard et al. 1985 ; Dwyer 1999 ).
Triangulation studies based on recall and other nonobjective measurements of
intake, for example, observations, weighed intakes and plate waste, indicate a
proneness to underreporting especially among individuals who are overweight and
the elderly (Beaton et al. 1979 ). Underreporting occurs with well-trained
researchers using standard protocols as in the US National Health and Nutrition
Examination Surveys (NHANES). Validity data collected periodically from 1971 to
2010 from a sample of more than 63,000 adults were used to calculate physio-
logically credible energy intake values based on estimated basal metabolic rate
(BMR) and estimated total energy expenditure.‘Disparity values’were computed
by subtracting these values from the NHANES-reported energy intakes (Archer
et al. 2013 ). The disparity values were−281 and−365 kilocalories per day for men
and women, respectively. The greatest mean disparity values were for obese
(BMI≥30 kg/m
2
) men at −716 kcal/day and women at −856 kcal/day. The
authors conclude that the reported intakes for 58.7% of men and 67.3% of women
were physiologically implausible and not compatible with long-term survival. Data
10 Objective and Subjective Aspects of the Drive to Eat in... 205