Conclusions and Implications: The serum vitamin D levels of subjects were low with small outdoor activities. In order to improve the current vitamin D status of Korean women, nutrition education programs should focus more on performing outdoor activities, especially during winter seasons, while maintaining adequate intake levels of vita- min D. Funding: NRF # 2010-0023838. P119 Gender Differences in Estimating Before and After a Meal David Sharp, BS, Deb246@cornell.edu, Cornell University, 341 MVR Hall, Division of Nutritional Sciences Mail Room, Ithaca, NY 14853; J. Sobal, PhD Objective: To study the effects of plate size on the accu- racy of meal size prediction and recall. Theory, Prior Research, Rationale: Prior quasi-exper- imental studies looking at perception of meal size have found differences in perceptions of meal size across plate size, with gender being a moderating factor. These studies were not done in the presence of an actual meal and could not validate if perceptions of meal size are accurate repre- sentations of actual meal size. Study Design, Setting, Participants, and Intervention: Quasi-experimental study of 300 college students eating lunch at a campus restaurant. Outcome, Measures and Analysis: Students were asked to draw either the meal that they had just purchased or the meal that they had just consumed on 1 of 2 sized plates; a plate that was the same size as was offered at the dining hall and one that was larger. Investigators recorded partic- ipant meals in order to compare drawn meal size with ac- tual meal size. Results: Drawn meals on same sized plates pre- and post- meal were accurate within 5% of actual meal size for both genders. Males given larger plates drew meals more than 10% larger than their actual foods. Females were more ac- curate than males when given larger plates. Conclusions and Implications: This study validates that students' drawings of meals can be accurate represen- tations of their meals and provides nutrition educators with a new tool for nutritional data gathering. It also sug- gests that conceptualizations of a meal can be influenced by size of the plate. Gender moderated accuracy of meal reporting. Funding: None. P120 Our Food Radius: Americans Purchase or Eat 75% of Their Food within Three Miles of Home Kathryn Hoy, PhD, RD, MFN, CD, foodandbrandlab@ cornell.edu, Cornell University, 31A Warren Hall, Ithaca, NY 14853; B. Wansink, PhD Objective: This research introduces the notion of our food radius: the area surrounding our homes that encom- passes most of our food purchases and consumption. Map- ping this area will underscore that challenging food issues may be global, but their solutions can be surprisingly local. Theory, Prior Research, Rationale: The psychological tendency of telescoping can distort problems – such as obe- sity – making them appear larger and less approachable than they actually are. Study Design, Setting, Participants, and Intervention: 53 adults who mapped out the distances to their most frequented grocery stores, restaurants, and workplaces from their home, and 502 women who were used as a validation sample. Outcome, Measures and Analysis: Measured distances of food shopping and eating locations from people’s homes, compared to their estimates. Results: The most frequented grocery stores and restau- rants are located within 2 miles of most people’s homes, and over 75% of all food is estimated to be either pur- chased or consumed within 3 miles of a person’s home. Conclusions and Implications: Changing one’s food environment can be done in a targeted, localized way. This research underscores that easy actions could be taken – very close to home – which can help individuals and their families eat healthier. Funding: None. P121 Visual Inspection of Cafeteria Waste by Quarter Servings Rivals the Accuracy of Weighing Kathryn Hoy, RD, MFN, CD, foodandbrandlab@cornell. edu, Cornell University, 31A Warren Hall, Ithaca, NY 14853; A. S. Hanks, PhD; D. R. Just, PhD; B. Wansink, PhD Objective: Weighing plate waste is the gold standard measure of determining what is eaten in school cafeterias, but it is costly and time consuming. Four common methods of analyzing tray waste – percentage estimates, quarter serving estimates, top-down photographs, and angled photographs – are compared to determine their accuracy. Theory, Prior Research, Rationale: No prior research has compared all of the leading plate waste estimation methods to determine which is most accurate and efficient to use in large scale operations. Study Design, Setting, Participants, and Intervention: Tray waste from 233 K-12 lunchroom trays in Lansing, New York were measured in each of 5 different ways: weighing, percentage estimates, quarter serving esti- mates, top-down photographs, and angled photographs. Outcome, Measures and Analysis: Estimation accu- racy compared with weighing. Results: Both methods of photography tended to under- estimate the amount of wasted food by approximately 12%. Percentage estimation showed a bias of 6% but had the lowest retest reliability. Quarter serving estimation (0, 25, 50, 75, or 100%) most accurately approximated actual weight (-3.2%). This is largely because most items are ei- ther in total, uneaten or only have 1to 2 bites (25%) out of them. P118 (continued) Journal of Nutrition Education and Behavior  Volume 45, Number 4S, 2013 Poster Abstracts S47 Continued on page S48