Proposal Phase II March 27, 2007 I. Variables 1. DV- presence of an eating disorder The main focus of the research is testing variables for their association with eating disorder presence. 2. IV- age Eating disorder sufferers are most often women and those of college age are the most affected. Thus, age is a key variable for eating disorder presence, so it will be important to know the age of the participants. 3. IV- self-esteem Development of an eating disorder requires that an individual have a negative body, which can be caused by factors such as insecurity and unrealistic body ideals. 4. IV- education level College-aged women are the most susceptible to <a href="/keyword/eating-disorders/" >eating disorders</a> . Yet, it is a necessary variable to determine if actually attending college or simply being of college-age is a relevant factor in eating disorder development. 5. IV- media exposure The media presents--and in some cases bombards--us with ideals for many aspects of our lives. Body image is one such ideal, and it is often unattainable and unrealistic for the average woman. Twenty-five percent of models can be diagnosed with anorexia nervosa in accord with the American Psychiatric Association (Hesse-Biber, Sharlene Nagy 3). 6. IV- sexism Those of the female sex are more likely to have <a href="/keyword/eating-disorders/" >eating disorders</a> , so the extent to which women agree with sexist concepts may have an effect on the development of <a href="/keyword/eating-disorders/" >eating disorders</a> . II. Conceptual Definitions 1. eating disorder- a disorder of the normal eating routine a. anorexia nervosa: a disorder usually occurring primarily in young women that is characterized by an abnormal fear of becoming obese, a distorted self-image, a persistent unwillingness to eat, and severe weight loss. It is often accompanied by excessive exercise, malnutrition, amenorrhea, and other physiological changes. b. bulimia nervosa: a disorder characterized by episodic binge eating that is followed by feelings of guilt, depression, and self-loathing. Bulimia is often associated with measures taken to prevent weight gain, such as selfinduced vomiting and the use of laxatives. 2. age- how many years the participant has been alive at the time of the survey 3. self-esteem- an individual's overall sense or evaluation of self-worth 4. education level- the number of years spent in school; also for this study, the current level of education (attending college, attended some college, or did not attend college) 5. media exposure- the extent to which the participant is immersed in popular culture through television and magazines or newspapers. 6. sexism- the extent to which the participant adheres to and believes in sexist gender roles prescribed by societal norms Proposal Phase II March 27, 2007 III. Measurement Procedures 1. presence of an eating disorder This is a nominal variable. It will be measured by completion of a test for the presence of an eating disorder and recorded as present or not present. I combined multiple tests and decided which responses qualified as an eating disorder. 2. age This is a nominal variable. It will be measured by participant's responses in whole number increments. 3. self-esteem This is an ordinal variable. It will be measured by completion of the Global Self-Esteem Scale. The data recorded will be the results of that test. 4. education level This is a two-part variable. The women in the experiment will identify two things: their highest education and their current status of education (attending college) or holding a job. The highest education will be presented as an ordinal variable and the current status of education will be presented as a nominal variable. 5. media exposure This is a ratio variable. Participants will answer (in whole-number increments) the number of hours they spend per week watching television and/or reading fashion magazines. 6. sexism This is an ordinal variable. I created an index and test to measure participant's level of sexism. The data will be recorded as low, medium, or high. IV. Tests of Validity and Reliability A. Validity: the extent to which measures indicate what they are intended to measure For this experiment to be considered valid, it should measure the presence of <a href="/keyword/eating-disorders/" >eating disorders</a> as affected by the five independent variables. 1. Face- The research question--Why are <a href="/keyword/eating-disorders/" >eating disorders</a> more prevalent in college-aged women?--is related to what I want to measure--the associations between the presence of an eating disorder and each of the independent variables. 2. Content- The goal of my research is to determine the effects my independent variables have on the presence of <a href="/keyword/eating-disorders/" >eating disorders</a> . Literature on the topic of <a href="/keyword/eating-disorders/" >eating disorders</a> suggests that age, sex, body image, and media exposure are key factors in predicting the presence of <a href="/keyword/eating-disorders/" >eating disorders</a> in individuals. Content validity is achieved by selecting independent variables that relate directly to these factors. 3. Criterion- This could be verified by having participants report their height and weight. The percentage of participants classified as having an eating disorder could be compared to the average percent of college-aged women who have <a href="/keyword/eating-disorders/" >eating disorders</a> . Proposal Phase II March 27, 2007 4. Construct- Discriminant validity is achieved when the independent variables do not influence one another: sexism, education level, age, self-esteem, and media exposure should not be highly influenced by each other. Convergent validity is achieved when multiple methods of assessing a construct produce similar results. Thus, the independent variables should generate comparable results in their determinations of the presence of <a href="/keyword/eating-disorders/" >eating disorders</a> . Reliability: a measurement procedure yields consistent scores when the phenomenon being measured is not changing B. 1. Test-Retest- To ensure participants would answer questions the same a second time as they did the first time, it is essential to provide clear instructions on how to answer the questions. Also, the questions must be easily understood and easy to answer. Moreover, participants should provide the same data as in the first test in any subsequent tests. 2. Internal Consistency- The tests that measure each independent variable should consist of questions that are highly associated with each other. Many of the surveys and indexes used in this experiment are from outside sources and are widely accepted. Random errors are less likely in tests that are easy for participants to understand and answer. 3. Interobserver- This experiment consists solely of personal responses from participants. Therefore, interobserver reliability is not a factor. V. Hypotheses 1. Women of college age are more likely to have <a href="/keyword/eating-disorders/" >eating disorders</a> than those of other ages. 2. Self-esteem and the presence of <a href="/keyword/eating-disorders/" >eating disorders</a> are negatively associated. 3. A woman's education level will not affect her likelihood of the presence of an eating disorder. 4. Media exposure and eating disorder presence are positively associated. 5. Sexism and the presence of <a href="/keyword/eating-disorders/" >eating disorders</a> are positively associated.