A matter of the metric? Sugar content overestimation is less pronounced in sugar cubes versus grams

Abstract

To make healthy food choices, consumers need to be aware of the sugar content of foods. Units act as an environmental cue that might influence sugar content estimation accuracy. The present study (1) tested whether estimations of sugar content are more accurate in sugar cubes vs grams, (2) compared accuracy of sugar content to estimations of the foods’ weight and energy content, and (3) investigated gender, education, and body mass index as potential correlates. A sample of 886 adults was randomly assigned to estimating the sugar content of 10 common foods in grams or cubes. Estimations of sugar content diverged considerably from actual values in both groups (0.22 ≤ Cohen's dsgrams ≤ 1.20; 0.20 ≤ Cohen's dscubes ≤ 1.10), but were more pronounced for sugar content estimations in grams in 7 out of 10 foods (ts ≥ 4.04, Ps < .001, Cohen's ds ≥ 0.14). Sugar content misestimation was somewhat more pronounced than misestimation of weight (0.05 ≤ Cohen's ds ≤ 1.43) and energy content (0.04 ≤ Cohen's ds ≤ 1.19). Relationships between sugar content misestimation and gender (0.00 ≤ Cohen's ds ≤ 0.33), education (–0.07 ≤ r ≤ 0.11), and body mass index (–0.08 ≤ r ≤ 0.06) were mostly negligible. Although sugar content estimations were somewhat more accurate in sugar cubes vs grams, estimation accuracy is generally low. In addition to promoting consumers’ knowledge through labeling and education, additional avenues for interventions might need to be explored for sizeable effects on food choices.