Young Children’s Figural Selections: Accuracy ofReporting and Body Size Dissatisfaction

Samantha Williamson* and Catherine Delin

School of Psychology, The University of South Australia, Adelaide, Australia

Accepted 12 May 2000

Abstract: Objective: The present study aimed to investigate children’s accuracy of reportingbody size and gender differences in body size dissatisfaction among preadolescent Australianchildren. Method: In an Australian partial replication of Dowdney, Woodward, Pickles, andSkuse (International Journal of Eating Disorders, 5, 29–40, 1995), children between 5 and 10years of age (N = 195) made current and ideal figure choices using the Body Image Percep-tion and Attitude Scale for Children (BIPAS-C), a nonsequentially ordered linear figural scale.Results: The results are consistent with accurate identification of their body size by Australianchildren at all the age levels. The girls, irrespective of age, preferred smaller ideal than currentshapes and expressed greater body size dissatisfaction than did the boys. Discussion: Theresults are consistent with the body dissatisfaction found in adolescent and adult women andwith those found in preadolescent children in the United States, Britain, and Israel. © 2000by John Wiley & Sons, Inc. Int J Eat Disord 29: 80–84, 2001.

Key words: children; body dissatisfaction; gender differences; figural ratings

INTRODUCTION

On the question of how accurately children perceive their body size, there are mixedfindings regarding gender differences in preadolescent children (Collins, 1991; Gustafson-Larson & Terry, 1992; Hill & Silver, 1995; Rolland, Farnill, & Griffiths, 1997; Tiggemann &Pennington, 1990; Tiggemann & Wilson-Barrett, 1998). All studies report that childrenregard their ideal shape as thinner than their current shape. However, it is not alwaysapparent that girls select smaller figures than do boys (Tiggemann & Pennington, 1990).Tiggemann and Pennington (1990) found no gender differences in body dissatisfaction inAustralian children 9–10 years of age. In contrast to Tiggemann and Pennington (1990),Wardle, Volz, and Golding (1995) and Collins (1991) found British and American girlsaged 4–11 years were more likely to choose thinner ideal rating figures than were boys.

*Correspondence to: Samantha Williamson, School of Psychology, The University of South Australia, Adelaide,South Australia 5000, Australia.

© 2000 by John Wiley & Sons, Inc.

Prod. #1589

The first aim of the present study is to clarify gender differences regarding body shape inan Australian sample of preadolescent children.

A technique with test-retest reliability and validity (Ben-Tovim & Walker, 1991; Collins,1991; Dowdney, Woodward, Pickles, & Skuse, 1995) frequently used to measure suchbody awareness in preadolescent children are figural stimuli, such as somatotypes, andsilhouettes (Collins, 1991; Rolland et al., 1997). Typically, subjects are asked to choose theircurrent and ideal body sizes. The discrepancy between these ratings is regarded as ameasure of body satisfaction (Tiggemann & Pennington, 1990).

The results of studies using figural rating scales to measure the accuracy of body shapereporting in children have varied. For example, different results are found when thefigural stimuli display is varied. Collins (1991) utilized seven male and female child andadult figures, ranging from very thin to obese in a fixed, sequentially ordered, linearfigural scale. She found children identified with the socially desirable middle-range (me-somorph) figure that was in the middle of the linear display. It was hypothesized that thechildren identified with the socially desirable, middle-range figure regardless of theirgender, weight, and body mass index (BMI).

In contrast to Collins (1991), Dowdney et al. (1995) incorporated a random, nonsequen-tially arranged figural scale to reduce the likelihood of children choosing the mesomorphfigure displayed in the middle range (as in Collins, 1991). Dowdney et al. (1995) foundthat the 10–11-year-old subjects chose figures consistent with their actual BMIs. As inCollins (1991), a direct comparison between BMI and figure selection was not made.Therefore, the present study had three aims: to clarify gender differences; to partiallyreplicate Dowdney et al.’s (1995) methodology in a group of Australian children; and todevelop an accuracy measure to ensure the appropriate exploration of Australian chil-dren’s body shape.

METHOD

Participants

One hundred ninety-five children participated, 94 males and 101 females. They rangedin age between 5 years 0 months and 10 years 8 months, with a mean age of 7 years and8.5 months (SD = 1.62). Subjects were sampled from five South Australian state-runschools from a range of socioeconomic neighborhoods in the Adelaide metropolitan area.The research was approved by the Human Research Ethics Committee at The Universityof South Australia, the Department for Education and Children Services (DECS), and theschool principals. A letter of introduction was given to the parents of each potentialsubject by the school and written consent was required from the parents before their childcould participate in the study. Consent forms were returned by 30% of the children. Thechildren were seen individually at their own school, but no names were recorded.

Apparatus

Height was measured, with shoes on and to the nearest 0.1 cm, with a height tape.Weight was measured to the nearest 0.5 kg with children in light summer uniforms. BMIcould thus be calculated. The first-named investigator took height and weight measure-ments after the interview procedure.

The Body Image Perception and Attitude Scale for Children (BIPAS-C) was adapted for

Young Children’s Figural Selections 81

use in the present study (Dowdney et al., 1995). Ten pictures of figures (five boys and fivegirls), ranging in size from very thin to fat with height held constant, were prepared onindividual laminated cards measuring 15 cm × 21 cm. Children were shown gender-relevant figures, otherwise the procedure was identical for all subjects. The nonsequentialsequence, that is, 2, 5, 3, 4, 1, as in Dowdney et al. (1995), was used on each test occasion.The children were asked to make the choices of ideal size, “Which picture would you liketo look like?” and current size, “Which picture looks like you?”

RESULTS

An ordinal scale of 5 points was constructed for BMI with reference to the ordered arrayof figural stimuli and percentile ratings recommended by the Australian Council onHealth, Physical Education and Recreation (Harvey & Althaus, 1993). BMI was groupedinto (1) underweight, (2) borderline underweight, (3) normal weight, (4) overweight, and(5) obese categories.

The accuracy of children’s figural selection was determined using absolute values bysubtracting the BMI ordinal scale score from children’s figural selection of their perceivedsize (current rating). Thus, a score of accuracy was calculated, which was then correlatedwith age using Spearman’s rank order correlations. There was no correlation, that is,accuracy did not increase with age. There was no significant difference between boys (M= 0.88, SD = 0.79) and girls (M = 0.82, SD = 0.78) on the accuracy of their figural selections.

In describing their current body size, girls selected a figure (M = 3.08, SD = 0.92) largerthan that selected by the boys (M = 2.97, SD = 0.96), although the results were notsignificant. In describing ideal self, an independent t test (t 193 = 5.42, p = .0001) showedthat the girls preferred a significantly smaller ideal figure (M = 2.31, SD = 0.88) than didthe boys (M = 2.97, SD = 0.82).

When assessing the discrepancy between ideal and current figure ratings, a paired t test(t 194 = 4.58, p = .0001) showed that the children selected a significantly smaller figure astheir ideal (M = 2.63, SD = 0.91) compared to their current figure (M = 3.03, SD = 0.94).However, as Table 1 highlights and as a t test (t 193 = 4.65, p = .0001) showed, this resultwas entirely accounted for by the difference between the girls’ current (M = 3.08, SD =0.92) and ideal (M = 2.31, SD = 0.88) figure preferences. As illustrated in Table 1, there wasno difference between the average males’ ideal (M = 2.97, SD = 0.82) and current figurepreferences (M = 2.97, SD = 0.96).

Table 1. Descriptive statistics

Variable

Male Female

M SD M SD

Age (years) 7.80 (N = 94) 1.67 7.89 (N = 101) 1.57Height (m) 1.29 (N = 94) 0.11 1.31 (N = 101) 0.17Weight (kg) 28.39 (N = 94) 7.21 29.64 (N = 101) 8.02BMI 16.79 (N = 94) 2.18 17.01 (N = 101) 2.67Child’s current body figure selections 2.97 (N = 94) 0.96 3.08 (N = 101) 0.92Child’s ideal body figure selections 2.97 (N = 94) 0.82 2.31 (N = 101) 0.88Difference between child’s

ideal and current body figure selections 0.00 (Median = 0.5) 1.09 −0.77 (Median = 1) 1.22(N = 94) (N = 101)

Note: BMI = body mass index.

82 Williamson and Delin

DISCUSSION

This study has partially replicated Dowdney et al.’s (1995) methodology in an Austra-lian sample of preadolescent children. As in Dowdney et al. (1995), the results of thepresent study are consistent with children at each age level (5–10) correctly identifyingtheir own body size. However, the present study has expanded on previous studies as anaccuracy score between subject’s BMI and current figure ratings was calculated to ensurethat a direct comparison between variables could be made.

The present study also clarified gender differences in body size dissatisfaction in asample of Australian preadolescent children. As predicted, children’s ideal figure pref-erences were thinner than their current figure choices. However, the result was entirelyaccounted for by the girls expressing dissatisfaction with their current body shapes. Thediscrepancy between ideal and current shapes was not found for the boys. Evidence forthe emergence of a slim ideal in Australian girls as young as 5 years of age was thusascertained.

Gender differences in body size dissatisfaction were also shown in children’s sponta-neous comments throughout the interview procedure. An example of a comment typicalof the girls was, “I weigh a lot!” In contrast, the boys made comments suggesting theywere satisfied with their body shapes, such as, “I’m happy with the way I am.”

Refinements are necessary when using figural stimuli to measure body image percep-tion. Ben-Tovim and Walker (1997) have suggested that use of perceptual tasks assumesthat people are able to access visual imagery and generate photographically accurate,pictoliteral, images. However, complex information processing, rather than simple match-ing procedures, appears necessary when comparisons are made between one’s own bodyimage and a pictorial presentation (Ben-Tovim & Walker, 1997). Gaining an understand-ing of children’s mental representations of their body image and of their ability to accessvisual imagery seems important for the applicability of linear figural scales and in theformation of appropriate body image measurement techniques.

We thank the principals, teachers, parents, and students from the participating schools for theircooperation and participation in the study.

REFERENCES

Ben-Tovim, D., & Walker, K. (1991). Women’s body attitudes: A review of measurement techniques. Interna-tional Journal of Eating Disorders, 10, 155–167.

Ben-Tovim, D., & Walker, K. (1997). Recall of the visual body image using a novel boundary detection task.Perceptual and Motor Skills, 85, 625–630.

Collins, M.E. (1991). Body figure perceptions and preferences among preadolescent children. InternationalJournal of Eating Disorders, 10, 199–208.

Dowdney, L., Woodward, L., Pickles, A., & Skuse, D. (1995). The body image perception and attitude scale forchildren: Reliability in growth retarded and community comparison subjects. International Journal of Meth-ods in Psychiatric Research, 5, 29–40.

Gustafson-Larson, A.M., & Terry, R.D. (1992). Weight-related behaviours and concerns of fourth-grade children.Journal of the American Dietetic Association, 92, 818–822.

Harvey, P.W.J., & Althaus, M. (1993). The distribution of body mass index in Australian children aged 7-15 years.Australian Journal of Nutrition and Dietetics, 50, 151–153.

Hill, A.J., & Silver, E.K. (1995). Fat, friendless and unhealthy: 9-year old children’s perception of body-shapestereotypes. International Journal of Obesity and Related Metabolic Disorders, 19, 423–430.

Rolland, K., Farnill, D., & Griffiths, R.A. (1997). Body figure perceptions and eating attitudes among Australianschoolchildren aged 8 to 12 years. International Journal of Eating Disorders, 21, 273–278.

Young Children’s Figural Selections 83

Stevens, C., & Tiggemann, M. (1998). Women’s body figure preferences across the life span. Journal of GeneticPsychology, 159, 94–102.

Tiggemann, M., & Pennington, B. (1990). The development of gender differences in body-size dissatisfaction.Australian Psychologist, 25, 306–313.

Tiggemann, M., & Wilson-Barrett, E. (1998). Children’s figure ratings: Relationship to self-esteem and negativestereotyping. International Journal of Eating Disorders, 23, 83–88.

Wardle, J., Volz, C., & Golding, C. (1995). Social variation in attitudes to obesity in children. International Journalof Obesity, 19, 562–569.

84 Williamson and Delin