Perceived and Individual Ideals of Gender in Swedish Adolescents with and without an Eating Disorder

Abstract

Three open-ended questions based on Steiner-Adair (1986) were used to study gender concepts in relation to ideals in adolescent boys and girls and eating-disordered girls. The answers were coded according to themes emanating from the collected data and not predetermined. Dichotomous differences were found between the genders regarding perceived ideals of body size, but clinical and nonclinical girls perceived ideals of society and media similarly. Individual gender concepts differed substantially from the perceived ideals, and were most often focused on lifestyle and life conditions. However, the clinical girls were somewhat more likely to emphasize appearance and loving relationships.

We are grateful to Hans Arinell for statistical and data processing support; Bruno Hägglöf, Carl Lago and Ulf Wallin for coding assistance and comments in the earlier steps of the development of the method of analysis. Thanks also to the participating students and to school staff who facilitated the assessments. The study was supported by Stiftelsen Sven Jerrings Foundation, the Swedish Research Council, Professor Bror Gadelius’ Memory Foundation and the Gillberg Foundation.

Notes

¹COEAT Research Protocol (1996): Coordinated research and evaluation programme for specialist units for treatment of patients with eating disorders in child and adolescent psychiatry in Sweden. Engström, I., Hägglöf, B., Hjalmers, I., Lago, C., Thurfjell, B. & Wallin, U. Unpublished manuscript.

²Data of the Three Questions of Ideals were available from eight boys, who gave responses comparable to the nonclinical group. However, the clinical sample is considered too small to draw any conclusions.

³“Characteristic theme” refers to a theme which is significantly more frequent in responses to one of the three questions compared to the other two questions.

⁴“Common theme” refers to a theme occurring in more than 10% of the sample.

**p < .01,

***p < .001,

††p < .01,

†††p < .001