Eating Disorder Inventory (EDI) is a standardized measure to evaluate eating behaviors and attitudes concerning eating disorders (Hersen & Beidel 2012). The measure is comprised of 64 self-report questionnaires which are organised into 8 categories. The questionnaires are based on a 6-point Likert Scale (Hersen & Beidel 2012). According to Baer and Blais (2010), ‘EDI is one of the most widely used measures for diagnostic impression, treatment planning, and outcomes evaluation with female eating disorder patients and non-clinical samples’ (p. 153). The EDI involves an ‘investigator-based’ approach because it is the investigator’s role to determine the behaviors and symptoms inherent in an individual (Hersen & Beidel 2012).
The measure has undergone a number of revisions, which have resulted in the inclusion of additional questions. This paper entails an evaluation of the efficacy of the inventory disorder inventory as a measure of assessing eating behavior and attitude. The analysis is focused on identifying possible areas that should be changed or improved. Additionally, the paper examines the reliability of the questionnaire applied in undertaking the EDI and how reliability can be improved.
Analysis of the EDI
Accurate and careful evaluation of eating disorders among individuals is critical in administering effective treatment and undertaking research. Goldman, Troisi, and Rexrode (2012) argue that the purpose of psychological assessment is to draw insight into an individual’s symptoms and hence develop an accurate diagnostic profile that can be relied on in making appropriate treatment recommendations. The EDI is very effective in developing a diagnostic profile. This arises from the fact that it aids in the identification of diverse clinical features, quantification of symptoms, and verification of diagnoses that can be relied on in addressing eating disorders (Hersen & Beidel 2012).
The effectiveness of EDI as a measure of analyzing the eating disorder is further enhanced by the fact that it is based on different subscales. The original EDI measure, which was comprised of an 8-subscale has been improved with the inclusion of additional subscales. The 8 subscales initially evaluated related to body dissatisfaction, drive for thinness, ineffectiveness, bulimia, perfectionism, interoceptive awareness, interpersonal trust, and maturity features (Clausen et al. 2010). However, the new measures, viz. EDI-2 and EDI-3 have additional facets. This has made the measure comprehensive in evaluating individuals’ eating disorders.
The EDI is based on individual reports, which are formulated using an extensive set of questionnaires (Goldman, Troisi & Rexrode 2012). Subsequently, it is possible to determine individual differences in eating disorders. This arises from the fact that the measure is applied independently (Hersen & Beidel 2012). Moreover, EDI can be used in assessing change in an individual’s eating behavior and attitude over time. The measure is not only efficient but also cost-effective.
Despite its effectiveness in identifying eating disorders among individuals, EDI cannot be applied in isolation. On the contrary, the efficacy of the measure in diagnosing and subsequently making recommendations on treating eating disorders among individuals is dependent on the degree to which clinical information is integrated (Goldman, Troisi & Rexrode 2012). Therefore, there is a need for stakeholders in the healthcare sector to ensure that a collaborative approach with clinicians is adopted to enhance the effectiveness with which EDI is applied in making decisions on treating eating disorders. This view is supported by Baer and Blais (2010) who affirm that ‘EDI cannot be used in isolation to make a diagnosis and must be interpreted along with clinical information derived from other sources of diagnosis’ (p.153).
In addition to this aspect, the effectiveness of the EDI in evaluating individual differences in eating disorders is significantly limited by the fact that the measure cannot successfully differentiate eating disorders amongst individuals arising from personal attitudes and behaviors from those that arise from personal psychological disorders (Gustafsson et al. 2010). This presents a significant gap associated with the measure.
Reliability of questionnaires
The development of self-reports, which are comprised of a set of questionnaires, is one of the methods that are commonly applied in assessing eating disorders (Goldman, Troisi & Rexrode 2012). However, for questionnaires to be effective it is imperative questionnaires to be characterized by a high degree of reliability, which constitutes one of the psychometric properties. Maruish and Moses (2013) define reliability as ‘the stability of an individual’s test scores over repeated administration of the test or alternate forms of a test’ (p.358). Findings of past research studies show that the questionnaires used in EDI are characterized by a high degree of reliability. One of the aspects that illustrate the reliability of the questionnaires entails the determination of score reliability, which is based on the concept of reliability generalization (Gleaves et al. 2014).
According to Gleaves et al. (2014), reliability generalization entails a method of meta-analysis that is applied in evaluating mean score reliability of a measure across studies to explore study factors that influence mean score reliability’ (p. 3). The questionnaires are characterized by a high degree of intern consistency. Weiner (2003) asserts that high test scores occur if there is a high degree of internal consistency, which depicts the reliability of the test. One of the measures used in assessing internal consistency reliability entails Cronbach’s coefficient Alpha, which evaluates the reliability of a particular multiple-item scale such as the questionnaires used in the EDI (Morgan & Griego 1998). In a study conducted by Gleaves et al. (2014) to measure eating disorder behaviors and attitudes, the findings revealed a considerably high Cronbach Alpha. The Cronbach Alpha of the EDI amounted to 0.91 for the total score while the score for the respective EDI subscale ranged between 0.75 and 0.89, which are substantially high.
Conclusion
The above analysis underlines that the Eating Disorder Inventory is an effective measure in diagnosing eating disorders among individuals. Nevertheless, some gaps should be addressed such as the inclusion of clinicians in its application and ensuring that it is effective in differentiating eating disorders as a result of behaviors and attitudes from disorders arising from psychological aspects. The analysis further shows that the questionnaires used in developing the self-reports are characterized by a high degree of reliability as evidenced by findings of high Cronbach co-efficient Alpha of studies conducted using the EDI questionnaires.
References
Clausen, L., Rosenvinge, J., Friborg, O., & Rokkedal, K. (2010). ‘Validating the Eating Disorder
Inventory-3; A comparison between 561 female eating disorders patients and 878 females from general population’, Journal of Psychopathological and Behavioural Assessment, 33 (1), 101-110.
Gleaves, D., Pearson, C., Ambwani, S., & Morey, L. (2014). ‘Measuring eating disorder attitudes and behaviors; a reliability generalization study’, Journal of Eating Disorders, 2 (6).
Goldman, M., Troisi, R., & Rexrode, K. (2012). Women and health. Oxford: Academic.
Gustafsson, S., Edlund, B., Kjellin, L., & Norring, C. (2010). ‘Characteristics measured by the Eating
Disorder Inventory for children at risk and protective factors for disordered eating in adolescent girls, International Journal of Women’s Health, 2 (1), 375-379.
Hersen, M., & Beidel, D. (2012). Adult psychopathology and diagnosis. Hoboken, NJ: John Wiley & Sons.
Maruish, M., & Moses, J. (2013). Clinical neuropsychology; theoretical foundations for practitioners.
Morgan, G., & Griego, O. (1998). Easy use and interpretation of SPSS for Windows; answering research questions with statistics.
Weiner, I. (2003). Handbook of psychology/ 10 assessment psychology. Hoboken, NJ: Wiley.
Ratings
