Young Adults’ Food Selection Patterns:Relations With Binge Eating and Restraint Janet A. Lydecker, Allison A. Palmberg, Katherine Vatalaro Hill, and Suzanne E. Mazzeo Binge eating is increasingly prevalent in college students (White, Reynolds-Malear, & Cordero, 2011). Binge episodes involve eating an objectively large quantity of food in a discrete amount of time and a perceived lost control over eating (American Psychiatric Association [APA], 2013). Strong negative affect commonly precedes and follows each episode. Binge eating is associated with severe obesogenic consequences such as diabetes, impaired heart function, and intestinal problems (Thompson, Roehrig, & Kinder, 2007). Individuals who binge eat report impaired functioning in personal, occupational, and social domains (Hudson, Hiripi, Pope, & Kessler, 2007). Addressing behaviors antecedent to a binge, such as food selection and acquisition, is an important part of understanding and creating interventions for this maladaptive behavior. Extant research has evaluated the selection and acquisition of food in laboratory settings (Raymond, Bartholome, Lee, Peterson, & Raatz, 2007), but fails to account for access to food within a specific community, such as a college campus. Context is important in understanding and treating binge eating because it influences perceived behavioral control of food selection. Universities provide ideal settings for the study of food selection because students have measurable options for food, a variety of choices within close proximity to where they live, and more equitable access to food options because of limited locations and dining plans, which have somewhat buffered prices. College is also a period during which young adults begin to make independent and long-lasting choices about what and how they eat (Cousineau, Goldstein, & Franko, 2004). Despite the critical developmental [End Page 493] period and unique environment, associations among students’ eating options, behaviors, and environment have not been evaluated together. Knowing how disordered eating relates to dining options will inform development of college-based interventions. The emotion regulation theory of binge eating (Chen, Matthews, Allen, Kuo, & Linehan, 2008) views disordered eating as a maladaptive attempt to cope with intense emotions, rather than a response to physiological hunger (Napolitano & Himes, 2011). Eating carries an expectation of escape from negative affect through the transfer of attention from the negative affect to a concrete object (i.e., food; Heatherton & Baumeister, 1991). Binge eating is a maladaptive coping strategy because it provides initial momentary relief, but ultimately negative affect (e.g., guilt and disgust) follows (APA, 2013). Before emotional eating can occur, students must (consciously or nonconsciously) decide what foods they will emotionally eat and must acquire the food. Awareness of food selection behavior can help college health providers assist students in identifying trigger foods and locations. This study contributes to existing research by assessing whether food selection differs in the presence of an emotion within the unique context of a college campus. METHOD Data were collected in two phases. First, 500 participants completed online surveys and received class credit for participating. Surveys measured disordered eating using the Three-Factor Eating Questionnaire (Mazzeo, Aggen, Anderson, Tozzi, & Bulik, 2003), a 36-item survey with three subscales: Restraint (TFEQ-R), Disinhibition (TFEQ-D), and Hunger (TFEQ-H). Surveys also asked where students ate both typically and when feeling emotions, using all locations accepting the institution’s dining plan. Second, 283 participants voluntarily completed a survey using an audience response system during class. These surveys also asked where students ate typically and when feeling strong emotions, but used categories of locations accepting the dining plan. Participants keyed answers using handheld devices called clickers. Clickers offer a nonjudgmental, nonthreatening format for self-report, because participants answer without providing identifiable data and large samples can be collected in a group context (LaBrie, Earleywine, Lamb, & Shelesky, 2006). Online survey respondents were students in psychology classes at a large Mid-Atlantic university (n = 500; Mage = 19.71). Clicker survey respondents were from large, nonpsychology classes at the same university (n = 283; Mage = 20.98). Online survey respondents were racially diverse: 22.8% African American (n = 114), 15.2% Asian American (n = 76), and 45.8% White (n = 229), largely representative of this urban, public university. Participants selected among ethnicity categories and had the option to use a write-in space. Clicker surveys did not collect ethnicity data to maintain...