Much of the literature on Disordered Eating Behaviors has examined their relationship with different risk factors. Among those factors examined have been perfectionism and stress. One form of perfectionism, Negative Perfectionism, involves the discrepancy or failure to meet high standards and order. There is also support for the relationship between Disordered Eating Behavior and Perceived Stress. One population presumably affected by Negative Perfectionism and Perceived Stress is graduate students. Yet, many of the existing studies on Disordered Eating Behaviors have either focused on or made primary use of individuals at the undergraduate-level and have neglected graduate student populations. Thus, the current study examined Disordered Eating Patterns within the graduate population, specifically focusing on Perceived Stress and Negative Perfectionism.

Participants were 108 female, graduate-level students, from two southern universities. The current study proposed that Negative Perfectionism, as measured by the Discrepancy Subscale of the Almost Perfect Scale-Revised (APS-R), would account for significant variation in Disordered Eating Behaviors, as measured by the three subscales (Uncontrolled Eating-UE, Cognitive Restraint-CR, and Emotional Eating-EE) of the Three-Factor Eating Questionnaire (TFEQ-R21), above and beyond variance accounted for by Perceived Stress, As measured by the Perceived Stress Scale (PSS-10). Also, the current study proposed that the PSS-10 would account for significant variation in Disordered Eating Behaviors, as measured by the three subscales (UE, CR, and EE) of the TFEQ-R21, above and beyond variance accounted for by Negative Perfectionism, as measured by the Discrepancy Subscale of the APS-R. Scores on the Discrepancy Subscale accounted for a significant amount of variance on the CR and EE Subscales, above and beyond the variance accounted for by scores on the PSS-10, but not on the UE Subscale. Scores on the PSS-10 predicted a significant amount of variance on the UE and EE Subscales, above and beyond the variance accounted for by the Discrepancy Subscale, but not on the CR Subscale.