Research Methods and Procedures:A total of 1069 fourth to sixth graders from nine elementary schools in the inner city of Philadelphia, PA, were part of an ongoing randomized control trial to assess prevention strategies for obesity. The mean rate of students eligible for free/reduced meals was 82.9 Ϯ 11.5%. Weight was measured in the second semester of the academic year. Absentee data for the entire academic year were recorded by homeroom teachers. Participants were classified into relative weight categories described by the Institute of Medicine: underweight, normalweight, overweight, and obese. Results: ANOVA showed that overweight children were absent significantly more than normal-weight children (12.2 Ϯ 11.7 days vs.10.1 Ϯ 10.5 days) (p Ͻ 0.05). Linear regression showed that the obese category remained a significant contributor to the number of days absent even after adjusting for age, race/ethnicity, and gender. Discussion: These data suggest that in addition to the medical and psychosocial consequences of being overweight, heavier children have greater risk for school absenteeism than their normal-weight peers. As the rate of childhood obesity increases, parallel increases in school absenteeism should be expected.

In a prospective longitudinal study, we examined whether the personality trait of self-control protects against weight gain during the transition from childhood to adolescence. We obtained multi-method, multi-source measures of self-control from a socioeconomically and ethnically diverse sample of 105 fifth-grade students. Height and weight were recorded by the school nurse and used to calculate age- and gender-specific standardized body mass index (BMI) z-scores. Self-controlled fifth graders had lower BMI z-scores in eighth grade compared to their more impulsive peers, and this relationship remained significant when controlling for potential confounds, including gender, age, socioeconomic status, ethnicity, IQ, and happiness. Moreover, when controlling for the same covariates, self-control measured in fifth grade predicted decreases in BMI z-scores from fifth to eighth grade. These results suggest that more self-controlled children are protected from weight gain in the transition to adolescence.

People seem to think that a unit of some entity (with certain constraints) is the appropriate and optimal amount. We refer to this heuristic as unit bias. We illustrate unit bias by demonstrating large effects of unit segmentation, a form of portion control, on food intake. Thus, people choose, and presumably eat, much greater weights of Tootsie Rolls and pretzels when offered a large as opposed to a small unit size (and given the option of taking as many units as they choose at no monetary cost). Additionally, they consume substantially more M&M's when the candies are offered with a large as opposed to a small spoon (again with no limits as to the number of spoonfuls to be taken). We propose that unit bias explains why small portion sizes are effective in controlling consumption; in some cases, people served small portions would simply eat additional portions if it were not for unit bias. We argue that unit bias is a general feature in human choice and discuss possible origins of this bias, including consumption norms.

The stigma-producing effects of "before and after" diet advertisements on a healthy weight sample were examined. Subjects (n = 59) were exposed to a presentation containing either a "before and after" diet ad, or solely the "before" or "after" picture embedded in a different ad. Subjects were then given measures to assess negative attitudes and endorsement of stereotypes about overweight people. Across all subjects, strong implicit anti-fat bias was present. Subjects in the Before and After condition indicated that weight is more easily controllable than did subjects in either the Before Picture Only or the After Picture Only conditions. There were two moderating variables for this effect. Subjects who reported greater life satisfaction endorsed fewer anti-fat stereotypes, and those who enjoyed an emotionally close relationship with an overweight person were less biased. These results suggest that "before and after" diet ads enhance weight stigma and perpetuate damaging stereotypes.

Segmenting a package effectively reduced consumption in the settings we have explored. Segmentation cues may operate by any or all of 3 mechanisms: (a) they call attention to and encourage better monitoring of eating, (b) they suggest smaller consumption (portion size) norms, or (c) they break automated eating sequences by introducing a pause. There is some evidence from the 1st study that provides evidence for the 1st, monitoring account.