Content area
Abstract
It has been well documented that the prevalence of anorexia nervosa has been increasing in both the United States and Europe (Willi, J., & Grossman, S. 1983). Recent studies have delineated several predominant personality variables which are characteristic of women with anorexia nervosa (Strober, 1986). Of particular interest to the present study is the characteristic of rigid and absolute thinking. To further explore the nature of rigid thinking patterns within anorexic women, this study compared twenty women with anorexia nervosa with a control group on measures of literalism (Landfield, 1980).
Literalism is a personal construct theory term defined as a way of feeling, valuing, and behaving which implies a restricted and rigid interpretation of the world. Literalism was examined within differing contexts and by comparing the two sub-types of literalism: literal-relating and literal-separating. Literal-relating refers to the excessive tightness of relationship between one's feelings, values and behaviors. Literal-separating refers to the excessive looseness between one's feelings, values and behaviors.
The procedure involved subjects completing the Eating Attitudes test (Garner & Garfinkel, 1979) and two modified forms of the Repertory grid (Kelly, 1955). The repertory grid is a sorting procedure used to elicit the personal constructs of individuals. The two repertory grids used were designed to elicit constructs within two different contexts: (1) self viewing other people and (2) situations involving food, weight and appearance issues. Using methodology developed by Leitner (1979), these constructs were differentiated into feeling, valuing and behavioral components and compared between and within groups.
Results indicated that the two groups did not significantly differ on measures of overall literalism. However, when the two different types of literalism were separately analyzed, the anorexic group was found to be significantly higher on the literal-separator score while the control group was found to be significantly higher on the literal-relator score. Results were inconclusive with respect to literalism within different contexts.
