Measuring Body Image Satisfaction: Digital Images vs. the Body Silhouette Method

Computer manipulation of actual digital photographs of patients has several advantages over the long-used and well-known drawing/silhouette method (Stunkard et al, 1983) for measuring body image dissatisfaction, according to two Australian researchers.

Drs. Rob Sands of Australian Catholic University, Fitzroy, Australia, and Christine Armatas of Deakin University, Victoria, Australia, report that their computer manipulation method has several advantages over the older method, including addressing the importance of dissatisfaction with specific areas of the body (J Psychol 2004;138:325).

The traditional method popularized by Stunkard and colleagues has been widely used in body image research, and consists of stylized and anonymous figure drawings representing physiques ranging from thin to obese. The individual selects the figure that most closely represents her actual self and then her ideal self. The difference between the two is used as a discrepancy score. This method has shown good test-retest reliability.

The authors point out that, in most studies the drawings are presented on a single sheet of paper and range in ascending size from left to right. On subsequent tests, individuals probably have little difficulty remembering which of the figures they marked before, particularly when the small number of figures is shown in ascending size. Others have pointed out that the figure drawing method has a restricted scale range and also that the figures depict people with obvious Caucasian characteristics, and thus don’t account for cultural and racial differences. (Gardner, R., in Body image, eating disorder and obesity in youth, Washington, DC 2001).

One of the major differences between the computer manipulation method used by the Australian team and the traditional silhouette method is that the computer program allows the clinician to measure the actual changes in the image made by the participant. These can be represented as a percentage of change at 5 body sites: the chest, waist, hips, thighs, and calves.

The comparison testing

The researchers compared both methods in a group of 56 female students 17 to 22 years of age. Participants were dressed in dark, tight-fitting clothing, such as leotards or a single-piece bathing suit, during the test. Each participant was then photographed digitally while standing in front of a blue background. A colored digital photograph was taken head-on. The digitized photographs were then downloaded into a laptop computer, cropped, and converted to an image in a frame. Participants then viewed two frames on the computer monitor—the original photograph could not be altered and acted as a reference point, while the second could be manipulated. The images were uploaded to a password-protected web site, where the students could log on and complete the test online. For privacy, the participants could only view their own image online, and the subjects were able to change the image at the 5 specific body sites.

Once online students were presented with 9 silhouettes and asked to indicate which represented their “actual” or “current” body shape and which represented their “ideal” body shape. Then they were presented with an image of their body that was 20% larger than their actual image. They were invited to indicate their “actual body shape and size” by re-sizing the image. Then they were presented with their actual body image and invited to indicate their “ideal” body size and shape by manipulating any or all of the five body sites.

What results showed

The mean BMI score for this sample was 22.1 kg/m2; 6 individuals were classified as overweight (BMI > 25) and 11 as underweight (BMI <20). On average, the participants selected an “ideal” figure/silhouette that was one to two sizes larger or smaller than their actual size. The mean for the average changes across the five body sites was –6.7. The mean for the sum of the percentage changes participants made to the actual image, to arrive at their ideal body size and shape, was –33.3%. Both these means indicate that for this sample their ideal was thinner than their actual size.

Advantages of the computerized system

But what is the advantage of a more time-consuming, equipment-based method such as the computer digital method as opposed to the effective and easy-to-administer silhouette scales? Unlike the silhouette method, the computer measurement protocol may evoke a more insightful, personal, and realistic appraisal of body image, according to the authors. They believe the computer manipulation method produces a sum of perceptual, cognitive, and affective responses to the five body sites, and these may differ in importance or intensity, or both. In addition, the researchers speculated that when subjects are confronted with an image of themselves, this might evoke different emotions and responses compared with the representation of line drawings of stylized images.