An evaluation of the construct of body image

Adolescence, Summer, 2002 by Sophie S. Banfield, Marita P. McCabe

Although early researchers conceptualized body image as being unidimensional, it is now considered to be, and is measured as, a multidimensional construct. However, the nature of these dimensions is unclear. Examples of dimensions are: perception, attitude, cognition, behavior, affect, fear of fatness, body distortion, body dissatisfaction, cognitive-behavioral investment, evaluation, preference for thinness, and restrictive eating (Brown, Cash, & Mikulka, 1990; Cash, 1994; Cash & Green, 1986; Cash & Henry, 1995; Gleaves, Williamson, Eberenz, Sebastian, & Barker, 1995; Slade, 1994; Williamson, 1990; Williamson, Cubic, & Gleaves, 1993; Williamson, Gleaves, Watkins, & Schlundt, 1993). The way in which body image is conceptualized is not just of theoretical interest, but has implications for the way in which disturbances in body image are treated.

The nature of the dimensions included in models of body image is diverse. Slade (1994) viewed body image as "a loose mental representation of body shape, size, and form which is influenced by a variety of historical, cultural and social, individual, and biological factors, which operate over varying time spans" (p. 302). Cash and his colleagues (Brown et al., 1990; Cash, 1994; Cash & Henry, 1995) viewed body image as being composed of perceptual and attitudinal dimensions. Gleaves et al. (1995) proposed a model that consisted of four dimensions: fear of fatness, body distortion, preference for thinness, and body dissatisfaction. These different models are reflected in measures of body image, making it difficult to compare body image findings using these different measures.

Empirical investigation is needed to determine the extent to which these models of body image accurately reflect the grouping of items. The aim of this research is to define body image more clearly and obtain data on the nature of the dimensions of body image. In doing so, it will be possible to develop a clearer idea of the actual dimensions of body image. Past conceptualizations of body image have generally incorporated at least one of the following four dimensions: perception, cognition, affect, and behavior. These four dimensions have the appealing features of being simple, functional, and clearly testable, and are the starting point for the model adopted in this paper.

Perceptual body image is defined as the accuracy of individuals' judgement of their size, shape, and weight relative to their actual proportions (Cash, Wood, Phelps, & Boyd, 1991; Slade, 1994). The study of perceptual body image involves assessing the accuracy of body size estimations, either at the level of individual body parts or the body as a whole (Cash et al., 1991). Two types of assessment procedures have been used for the measurement of perceptual body image: paper and pencil format, and body image accuracy techniques. However, the fundamental methodological problem with measures that use a paper and pencil format is that they fail to assess individuals' actual body size and therefore have no physical reference point to compare to individuals' judgement of their body size. As a result, there is no reference point or physical measurement by which to determine individuals' perceptual distortion.

Body image accuracy techniques include two types of assessment procedures-size estimation techniques (e.g., the Movable Calliper Method; Reitman & Cleveland, 1964; Slade & Russell, 1973) and distortion techniques (e.g., Askevold, 1975; Brodie, Slade, & Rose, 1989). The distorting image techniques are used more frequently, and these include the distorting mirror (Traub & Orbach, 1964), distorting photographs (Gluckman & Hirsch, 1968) and distorting video camera (Free man, Thomas, Solyon, & Hunter, 1984) techniques. These methods involve subjects adjusting an image of themselves until it corresponds with how they perceive their body (Brodie et al., 1989). The problem with these images is that adjustments work on the body as a whole, and it is difficult to alter the size of one part of the body independent of other parts (Monteath & McCabe, 1997).

The second two aspects of body image are the affective and cognitive dimensions. The affective dimension can be conceptualized as the feelings individuals have towards their bodies' appearance (Cash & Green, 1986). The cognitive component relates to thoughts and beliefs concerning body shape and appearance (Cash & Green, 1986). Often measures that claim to be measuring one of these components exclusively actually contain items that relate to the other dimension, or include items relating to both dimensions in the single measure. For example, items in the Body Dissatisfaction Subscale of the Eating Disorder Inventory (Garner, Olmstead, & Polivy, 1983) assess both thoughts and feelings (e.g., "I feel satisfied with the shape of my body" and "I think my hips are too big").

Support for the partitioning of the affective versus cognitive aspects of body image has come from Thompson and Psaltis (1988), who found that individuals have both a rational and affective view of their body size. Thompson and Psaltis, after modification of the Fallon and Rozin (1985) protocol, found that figural ratings were higher when respondents were instructed to base their ratings on how they feel (affective instruction), rather than what they think (cognitive instruction), about their appearance. This corresponds to other research findings (Bowden, Touyz, Rodriguez, Hensley, & Beumont, 1989; Franzen, Florin, Schneider, & Meier, 1988; Proctor & Morley, 1986).