This item is only available as the following downloads:
Running Head: DRAWING TASKS FOR BODY PERCEPTION Using Drawing Tasks to Study Body Perception Andrew Fishman A Thesis Submitted to the D ivision of Social Sciences New College of Florida in partial fulfillment of the requirements for the degree Bachelor of Arts Under the sponsorship of Dr. Heidi Harley Sarasota, Florida May 2012
DRAWING TASKS FOR BODY PERCEPTION ii Acknowledg ments I wish to give my sincerest thanks to my family, friends peers, my baccalaureate committee, and Dr. Heidi Harley, without whose emotional thesis would not have been possible.
USING DRAWING TASKS TO STUDY iii Contents Abstract ................... ....................................................... ...............................v ....................................................... ...... .... ..2 Surveying Body Image and Perceived Attractiveness ........................ .. .... ..6 Visualizing and Manipulating Individual Body Dimensions ................. ..8 Drawing Tasks .... ............................................................... .......................... .11 Current Study ..................... ......................................................................... 1 5 Method ..................................................................................................... .... 17 Results ... .................................................................................. ..................... 20 Discussion ................................................................................................. ... 2 4 References ... ....................... .......................................................................... 26
USING DRAWING TASKS TO STUDY iv List of Figures and Tables Appendix Eating Disturbance Scale (EDS 5) ............................................................. 3 1 Figure 1 ..............................3 2 Figure 2 Average Body Perception Index for drawings o f Self and Other..............3 3 Figure 3 Average Error for drawings of Self and Other............. ..............................3 4 Figure 4 .........3 5 Figure 5 Self Drawn Hip Dimensions compared to Body Mass Index....................3 6
USING DRAWING TASKS TO STUDY v Abstract Using Drawing Tasks to Study Body Perception Andrew Fishman New College of Florida, 2012 ABSTRACT correlates with body ima ge and BMI, but it lacks the necessary subtlety to assess individual body dimensions Other methods for assessing body image include surveys and various approaches to estimate body measurements. Previous research has suggested that people are less accurate in perceiving their own Drawing tasks have been useful for studying psychological phenomena and may be a sensitive measure for studying body image. In this study, p articipants drew themselves and a friend Figure Rating Scale reported measurements to determine Body Mass Index, and completed a survey on risk of eating disorders. Participants drew the lower halves of thei r bodies (waist and hips) more accurately than their upper halves (shoulders and chest). No significant differences were found between perceptual accuracy of the drawings of Self and Other. The s e finding s could be due to the homogen eit y of the group, to to participants drawing a standardized human figure, or to participants drawing themselves for both drawings Dr. Heidi Harley Division of Social Science s
DRAWING TASKS FOR BODY PERCEPTION 1 Americans spend a considerable amount of time and money focused on their looks, and they are often unhappy with the results. A Psychology Today survey in 1997 found that 56% of females and 43% of males who responded to a self image questionnaire reported dissatisfaction with their overall appearance (Garner, 1997) In extreme cases, this dissatisfaction can become clinical and result in di sorders such as Body Dysmorphic Disorder, Anorexia Nervosa and Bulimia Nervosa. Body Dysmorphic Disorder, also referred to as Body Dysmorphia, refers to a disorder in which people have an obsession with a minor or imagined flaw in their physical appearan ce. Body Dysmorphia is currently diagnosed based upon three criteria: One must have an excessive preoccupation with a minor or imag appearance; this preoccupation must cause significant distress or im everyday life; an d the symptoms must be distinct from those of another mental disorder, such as Anorexia Nervosa or Bulimia Nervosa (American Psychiatric Association [DSM IV], 2000) Anorexia Nervosa is similar, but differs from Body Dysmorphia in that the preoccupation with physical appearance manifests primarily in a focus on In order to be diagnosed as Anorexic, a person must refuse to maintain normal body weight, fear gaining weight, place excessive importance on weight for self worth, and, for postmena rcheal females, experience the absence of three consecutive menstrual cycles (American Psychiatric Association [DSM IV], 2000) Bulimia Nervosa is similar to, but distinct from Anorexia Nervosa Like people with Anorexia Nervosa, people with Bulimia Ner vosa focus their self worth on body shape and weight. People with Bulimia Nervosa, however, b inge on food, eating more than they physically desire (often feeling like they cannot control their eating behavior
DRAWING TASKS FOR BODY PERCEPTION 2 during binging episodes), and then purging the food from their system, most commonly through self induced vomiting (American Psychiatric Association [DSM IV], 2000) P eople with eating disorders misperceive their own bodies in various ways ( Calden, Lundy, & Schlafer, 1959; Channon, de Silva, Hemsley & Mukherjee, 1990 ; Fallon & Rozin, 1985 ; Sand, Lask, Hie, & Stormark 2011 ; Scagliusi, Alvarenga, Polacow, Cords, de Oliveira, Coelho, & Lanchal, 2006 ; Slade & Russell, 1973 ; Thompson & Altabe, 1991 ) Finding out exactly how a person sees him/herself i s important, not only in determining treatment, but also so that therapists can help their patients move towards accurate self perception s The current study will introduce a new method for studying body image, i.e., Self and Other portraiture in order to determine its usefulness in analyzing misperceptions of specific body areas. Before the study itself, I will review other methods of assessing body image and some of the studies in which they have been adopted. Figure Rating Scale (FRS) was created in 1983 in order to assess perception of weight status of participants (Stunkard, Sorensen, & Schulsinger, 1983) It consists of 9 male figures and 9 female figures arranged from very thin to very obese (Figure 1). Participants are traditionally asked to indicate which figure they feel is the most commonly used measure for assessing body size awareness. The figures participants are asked to estimate on the scale occasionally vary depending indicate 3 figures on the scale rather than 2: one figure that represents a n represents which figure best represents how others see the participant.
DRAWING TASKS FOR BODY PERCEPTION 3 Thompson and Altabe (1991) between figure most a ttractive to the opposite sex. Researchers in this study asked participants to select six figures from sented ventory (EDI) to assess eating disorder behavior. Two groups of students were studied for different purposes. A group of 92 students (58 female, 34 male) were evaluated twice at a 2 week interval to test retest reliability, and 237 students (146 female, 91 male) were evaluated once to test for validity. The difference between figures chosen for how participants reported feeling and how they thought they looked was positively associated with greater eating disturbance and was negatively associated with lo wer self esteem. Researchers concluded that has good test retest reliability and validity and that the way people think they look and the way they feel are different and should be treated as such. Non Western populations can also be assessed accurately through use of Stunkard (Lo, Ho, Wong, Mak, & Lam, 2011). Adolescent students (N=5666) in reported their height, weight, perceived weight status, and health related quality of life (assessed using the Medical Outcomes Study Short Form version 2). Researchers also measured height and weight objectively. Current body size, as measured by figure s FRS was strongly correlated with BMI (both reported and measured) but was only weakly associated with health.
DRAWING TASKS FOR BODY PERCEPTION 4 Body perception as measured by figure rating can predict future weight loss ( Lynch, Liu, Wei, Spring Kiefe, Greenland, 200 8). In a longitudinal study conducted with 3,665 participants over a period of 13 years, participants rated current self body size and ideal body size as well as several demographic questions. Researchers used the difference between t hese two body size for each participant, once at baseline and once 13 years later. Obese women who rated themselves as obese (figures 8 and 9) gained significantly less weight over the 13 years than obese women who perceived themselves as normal weight (figures 3 and 4). Similarly, obese women who had a body size satisfaction of 1 (perceiving their body as a body size satisfaction themselves as obese and wish they were thinner gain less weight over time. also be measured by rating figures on a similar scale (Channon, de Silva, Hemsley, & Mukherjee, 1990). Forty eight female anorexic patients and 24 female controls were asked to rate, from a series of 6 line drawings of a female over her recovery from Anorexia (from very thin to average ), how they perceive d themselves, how others perceive d them, and how they desire d to be. Both anorexic and control females perceived themselves as larger than their ideal weight. Anorexic patients who had maintained a low weight during the recovery process tended to prefer a slightly larger size picture, while anorexic patients who had gained weight in the recovery process preferred pictures smaller than they perceived themselves to be. The difference between by psychologists who were working separately with the patients.
DRAWING TASKS FOR BODY PERCEPTION 5 Stunkard Figure Rating Scale was used to distinguish between female students with and without Bulimia Nervosa in Brazilian populations (Scagliu si, Alvarenga, Polacow, Cords, de Oliveira, Coelho, & Lanchal, 2006). Participants (N=114: 98 control, 16 diagnosed with Bulimia Nervosa) selected two figures on the scale, one that represented their current body size and a second that represented their ideal body size. with diagnosed Bulimia Nervosa, indicating that these self selected sc ores could be a valid measure of body image. Neither males nor females accurately estimate the body type of the most attractive person of their own gender (Fallon & Rozin, 1985). Two hundred forty eight male and 227 female undergraduates used Stunkard Figure Rating Scale to indicate four figures: one figure that represented their current size a second figure that represented their ideal body size, a third figure they perceived as the most attractive figure to the opposite sex, and a fourth figure they would find most attractive in the opposite sex. For men, there was no significant difference among all three of the male figures, but for females, their heavier than most attractive was compared to the figure they selected as the one they thought the other attract females underestimated it. only figure rating scale in use Calden, Lundy, & Schlafer (1959) used another figure rating scale to sh ow that b oth males and females prefer a balanced body type, although each gender misperceives what the opposite gender
DRAWING TASKS FOR BODY PERCEPTION 6 prefers. Females find thinner males more attractive than males believe, and males find balanced females more attractive than females bel ieve. One hundred ninety six female and 110 male students filled out a self concept questionnaire that included questions relating to their satisfaction with specific areas of their body, such as height, waist, and chin size. Each participant was also sh own a scale of seven figures of each sex from endomorph to mesomorph, i.e ., from overweight to muscular. The females reported wishing to be smaller in every area except for bust size and reported the most d issatisfaction with their body from the waist down Males reported most dissatisfaction with their body from the waist up However, both sexes reported that they found the bala nced figure the most attractive; females viewed the thinner male as more attractive and less attractive than the males did. T hese data suggest that people focus on specific aspects of their bodies when they analyze them. Surveying Body Image and Perceived Attractiveness Some researchers have attempted to address the de pth of the issue by directly asking participants about their feelings and behaviors. Participants in these studies often fill out several surveys in order to determine which of their attitudes or characteristics are related to unhealthy behavior. Self e steem and body dissatisfaction are related to muscle dysmo rphia in men (Ey, 2010). One hundred one male gym goers were surveyed on muscle dysmorphic endorsement (belief that one is too small and should gain muscle mass), self esteem, and body dissatisfac tion. A ll three were related to one another such that as muscle dysmorphia increased, self esteem decreased and body dissatisfaction increased Heterosexual men tend to have a more positive self image than homosexual men, homosexual women, and hetero sexual women (Peplau, Frederick, Yee, Maisel, Lever &
DRAWING TASKS FOR BODY PERCEPTION 7 Ghavami, 2009). Thousands of lesbian women, gay men, heterosexual women, and heterosexual men were polled online about comfort with their bodies during sexual activity, comfort with wearing a swimsuit in public, and preoccupation with weight. Heterosexual men reported the fewest concerns in all areas, but there were no differences among any other groups. In 2006, Lev esque & Vichesky surveyed 64 gay men about body satisfaction, perceived acceptance within the gay community, soc ial comparison tendencies, body image satisfaction, self esteem, and depression. These men tend ed to be more concerned with muscularity than weight, and body perception was negatively associated with BMI and a tendency to compare themselves to others. Brown and Graham (2008) sur veyed 77 males, 38 heterosexual and 39 homosexual in order to examine body satisfaction and reason for exercise for men of different sexual orientation. Participants filled out an ar ray of tests designed to test masculinity, femininity, narcissism, and reason for exercise. Heterosexual males were more likely to exercise for fun than homosexual males, and homosexual males were more likely to exercise to improve their appearance. Masc ulinity rating was also positively correlated with body satisfaction. Homosexual males who scored low on masculinity were least satisfied with their bodies; heterosexual males who scored high on masculinity were most satisfied. For both homosexual and bisexual men, gym behavior (active gym membership, frequency of visits to the gym, length of visits at the gym, etc.) is strongly predictive of body dissatisfaction and eating disorders (Siconolfi, Halkitis, Allomong & Burton, 2009). Gay men bisexual men and other m en who have sex with men (MSM) ( heterosexual people, females, and t ransgender people were excluded) were surveyed for risk factors in eating disorders, including frequency of exercise session, anxiety, concern about external
DRAWING TASKS FOR BODY PERCEPTION 8 perception, HIV st atus, and length of residence in New York City. Body dissatisfaction was related to age, but unrelated to race and HIV status. Sixty percent of the gay, bisexual, and MSM men surveyed had an active gym membership and worked out an average of 3 days per w eek. Eating disorder scores were negatively correlated with hegemonic masculinity, suggesting that men who do not believe men need to be strong and dominant are more likely to have eating disorders. This can be explained by previous research suggesting t hat eating disorders are a feminine pathology, or that childhood non conformity is related to body dissatisfaction. Visualizing and Manipulating Specific Body Dimensions In order to determine how people view themselves, some researchers have created alternative methods that let participants manipulate tangible tools that represent body measurements. Some include physically estimating the size of individual body parts, and wn. Patients suffering from Anorexia Nervosa overestimate the width of their own bodies but not the width of other objects or models or height in general (Slade & Russell, 1973). In one study, 14 participants (13 female, 1 male) estimated their own widt h at various points (face, chest, waist, and hips) using two lights placed on a horizontal track bodies, body image perception index was determined by calculating (per ceived size) / (real size) x 100. Control participants were accurate in estimating their measurements, but patients with Anorexia Nervosa overestimated their widths from 27 58%. A second study by the same researchers sought to determine whether this dis only applicable to the patients themselves. Ten patients (9 female, 1 male) receiving care for Anorexia Nervosa completed a task similar to the first described above, but they were
DRAWING TASKS FOR BODY PERCEPTION 9 also asked to estimate the various widths of a model, as well as the height of both themselves and the model, using a vertical version of the measurement device. Patients with Anorexia Nervosa overestimated their own width more than they did fo r the model, but they were much more accurate in assessing the heights of both the model and themselves. their bodies, i.e. their width versus their height. The third study in this series measured what happened to body perception as the patients gradually gained weight. To do this, the first study was replicated weekly for the course of Anorexia Nervosa treatment until discharge. As patients gained weight, t hey became more accurate in perception of their width. Because they were in treatment, this finding was interpreted as an indicator of better health. Another study investigated the same overestimation in a different way (Sand, Lask, Hie, & Stormark 2 011). Several distorted photographs were presented to 406 adolescent participants (240 female, 166 male): one photograph of themselves, two of peers (a male and a female), a neutral object (a milk carton), and a female research assistant. Each photograph was stretched either 50% or 150% of original width, with a randomized (up to 5%) additional distortion. Participants then used a computer program to stretch or shrink the picture back to where they perceived it as accurate. Estimation accuracy was measu picture type, with positive scores reflecting overestimation and negative scores reflecting thinness, emotional instability, self esteem, overeating, and maturity fears, to test for eating disorder symptoms. Participants with higher scores on the survey were considered to be at higher risk of eating problems. Both genders estimated the width of the milk carton accurately. Females with higher scores were more likely to overestimate their own
DRAWING TASKS FOR BODY PERCEPTION 10 width and underestimate the width of other females. In contrast, males were likely to underestimate their own width and overestimate the width of other males. Fa thers of female patients with eating disorders also wish for changes in their own bodies, but they perceive their current state correctly (Benninghoven, Tetsh, Kunzendort & Jantschek, 2007). Forty two patients and their fathers created 2 computer gene rate d models of their own body: perceived and ideal ; the fathers also created models of their daughters. The fathers generally perce ived their own bodies correctly but wished to be thinner and more muscular. Desire for change in body fat was strongly correla ted between daughter father pairs for both Anorexia Nervosa and Bulimia Nervosa patients. physical estimations of body size, and various surveys. These are successful, but f or many people, body dissatisfaction is more specific than these studies indicate. One might s way, one the torso. Drawing tasks also require easily accessible technology i n that almost everyone can access paper, pens, and rulers. Drawing Tasks Drawing tasks have often been used for psychological assessment. For example, in 1926, Dr. Florence Goodenough intelligence lev els and mental ages. She score d drawings by tallying how many of 51 distinct criteria were met In order to test the validity of her test, Goodenough administered 334 children the test along with the Stanford Binet. Results were significantly correlated with one another, r = .44.
DRAWING TASKS FOR BODY PERCEPTION 11 In 1948, John Buck introduced Tree assess psychological well being in children. In order to administer the test, one asks the child to draw a house, then a tree, and then a person. The drawings are assessed on a variety of criteria to assess psychological health. This test is still used. Milne and Greenway (1999) tested 61 children from ages 4 14 who drew five drawings : the House Tree not diminish as they got older as did the boys Researchers inferred that males tend to suppress emotions as they reach puberty, and this could manifest itself in their use of colors. Y oung and old males differed in their use of color between drawings, and it The HTP test was not useful in measuring psychological self es teem (Groth Marnat, 1998). Forty graduate students were given the House Tree Person test as well as the Human Figure Drawing test, which asks the participant to draw a person, then another of the opposite sex; the Coopersmith Self Esteem Inventory; and t he Tennessee Self Concept Scale. The participants were also asked to rate their own drawing ability on a 5 point scale. The researchers evaluated the drawings quantitatively using 5 point strong appearance of scores on the drawings were not significantly related to the other standard measures of self esteem. Human figure drawings can be used t persona lities (Koppitz, 1966). In a child guidance clinic, 31 male and female students children who were ch child individually completed a human f igure d rawing task. The drawings were graded on 30 variables, including poor
DRAWING TASKS FOR BODY PERCEPTION 12 integration of parts, gross (large) asymmetry of limbs, crossed eyes, hands cut off, genitals, and three or more figures spontaneously dr awn. The aggressive children tended to include gross asymmetry of limbs, teeth, long arms, big hands, and genitals. The shy child ren tended to draw tiny figures mouth, and hands. Obese males tend to draw l arger figures than normal weight males and obese females when unprompted (Shenker, 1977). This study was conducted to examine self perception in normal weight vs. obese individuals. Shenker used the Draw a Person (DAP) test, in which participants are ask ed to draw a person, after which they are prompted to draw a person of the opposite sex as the first drawing. The first drawing is usually considered by experimenters using this test to be a self portrait, even though the person may not be consciously awa re of it. Experimenters rated the drawings on several criteria, such as size and presumed accuracy. They found that the drawings completed by the normal weight participants ll as The Draw A Person (DAP) test has been used in multiple ways. Teglasi (1980) used it to determine acceptance of traditional female roles (Teglasi, 1980). One hundred fifty college women half of whom were members of the feminist group National Organizatio n of Women (NOW) we re given the Draw A Person Test Women who belonged to NOW for their first person in the DAP Test. Twenty six percent of non members drew males
DRAWING TASKS FOR BODY PERCEPTION 13 for their first person. Researchers concluded that this was due to differences in acceptance of traditional gender roles. The Draw A Person: Screening Procedure for Emotional Disturbance (DAP:SPED) can be used to help assess children with emotional disturbances (Matto, 2002). The test was given to 68 children from 6 12 years old who were receiving counseling services. The DAP:SPED requires the participant to draw a man, then a woman, then him/herself. The test was modified from the DAP test to help identify children and adolescents with emotional or behavioral disturbances. The drawings are rated on 55 items, each scored either 0 or 1, notating absence or presence of the particular item. The items were totaled and compared to a shortened version of the Child Behavio r Checklist and the Child and Adolescent Adjustment Profile. Measures were significantly correlated, indicating that this test may be used to help identify children with emotional disturbances. The Draw A Person test was not useful in diagnosing conduc t disordered children (Feyh & Holmes, 1994). Eighty children, aged 10 16, completed the Draw A Person test. Each group of 40 had 20 females and 20 males. Half (40) had histories of conduct disorders icators of like or grotesque figures, clenched fists, omission of arms, talon like fingers, visible teeth, slash line mouth, and toes on a clothed ders or groups. Figure drawing tasks can be used to assess self concept and family functioning (T haringer & Stark, 1990). Fifty two children were given the DAP test as well as the Invento A ffective Disorders and Schizophrenia for School age Children, the Coopersmith Self Esteem
DRAWING TASKS FOR BODY PERCEPTION 14 Inventory, and the Self Report Measure of Family Functioning. This battery thoroughly tested the children for symptoms of depression, anxiety, and family issues, which were which s The KFD drawings were scored twice first, us ing the Reynolds KFD System, which consists of 37 indicators, scor ed for presence or absence, and second using a qualitative system that measured Psychological Functioning of the Family on a scale from 1 to 5 1 being the absence of psychopathology, and 5 being the presence of severe psychopathology. Scores from each test were significantly correlated with self reported self concept and family functioning. Researchers concluded that such projective drawing tasks can be useful in assessing children with s ome disorders. Current Study bodies. People suffering from Anorexia Nervosa overestimate the width of t heir own bodies but not the bodies of other people or objects (Channon et al., 1990; Sand et al., 2011). Patients with Anorexia Nervosa within clinics improve in accuracy on proportion estimating tasks as they progress in a mental health clinic ( Channon et al., 1990; Slade & Russell, 1973 ). Sexual orientation affects what one finds attractive, both for oneself and balanced body type in the opposite gender, but males overestimate the width of the most attractive male body and females unde restimate the width of the most attractive female body (Calden et al., 1959; Fallon & Rozin, 1985). For males, this is related to level of
DRAWING TASKS FOR BODY PERCEPTION 15 muscle dysmorphia (Ey, 2010). Heterosexual males have more positive self images than homosexual men, homosexual wom en, and heterosexual women (Peplau et al., 2009). Homosexual men are more concerned with muscularity than they are with weight (Levesque & Vichesky, 2006). is very effective at assessing current and ideal body size, but it lacks nuance. People who wish they were different in one area but not in others may find it impossible to choose a Stunkard figure that accurately represents their current or ideal selves. Channon et al. (1990) emphasized the importance of this problem when they found that men tend ed to be less satisfied with the upper half of their bodies and women were less satisfied with the lower half. Women also tended to want smaller dimensions in every area except for breast size (Calden, 1959) There are not e nough figures on the Stunkard scale for those women to express that, because the figures with larger breasts are also larger in other dimensions. Other measurement estimation tasks have been successful for body parts, but they require technology that may be inaccessible to most therapists ( Benninghoven, Tetsh, Kunzendort & Jantschek, 2007 ; Sand, Lask, Hie, & Stormark 2011 ; Slade & Russell, 1973 ). A drawing task may deliver this nuance via readily available materials. Drawing tasks in psychology have be en shown to relate to intelligence in children (Goodenough, 1926), gender (Deaver, 2009; Milne & Greenway, 1999), level of maturity (Milne & Greenway, 1999), self concept and family functioning (Benziger, 1982; Tharinger & Stark, 1990), aggression in child ren (Koppitz, 1966), acceptance of traditional female roles (Teglasi, 1980), emotional disturbances (Matto, 2002), and sexual abuse (Tillman 2004 ). Such tasks may be sensitive measures for studying body image too.
DRAWING TASKS FOR BODY PERCEPTION 16 The current study seeks to expand on t he research on body image focusing on the subtleties that are overlooked by traditional figure scales and using accessible technology By drawing themselves, participants may be able to report more clearly how they feel their bodies look and how they would like them to look In this study, p proportions will be compared to photographs of the participants to test whether people, especially those at higher risk for bo dy dysmorphia, are inaccurate about their own body perception in the same way that they are with their own self image. The study will also test whether this can be extended to different genders and sexual orientations, as previous research has suggested.
DRAWING TASKS FOR BODY PERCEPTION 17 Pr edict ions : 1 Participants will be more accurate when drawing other s than they are when drawing themselves. 2 drawings will manifest in over estimation of specific drawings wi ll manifest in under estimation of specific body dimensions. 3 Heterosexual men will have the fewest inaccuracies when depicting themselves. 4 Overall inaccuracy in self perception will be related to risk of eating disorders. 5 Drawn body dimensions will be related to Body Mass Index ( BMI ) Figure Method Participants Eighteen p articipants (15 female, 2 male, 1 other) attending a small liberal arts college in southwest Florida volunteered to participate in the study Fifteen of the 18 part icipants reported being heterosexual. Particip ants ranged in age from 18 to 22 Participants were recruited via email Materials variety of colored pens were provided to each participant. Each parti cipant filled out the Eating Disturbance Scale (EDS 5) (Rosenvinge, Perry, Bjrgum, Bergersen, Silvera & Holte, 2001) ( Appendix ) Each participant also took Stunkard Scale (Stunkard, Sorensen, & Schulsinger 1983) (Figure 1) to measure body perception.
DRAWING TASKS FOR BODY PERCEPTION 18 Procedure Participants paired themselves with a partner to whom they felt close. This was defined by the individual participants. Each participant and his/her partn er were assessed simultaneously but independently from other pairs. This assessment was done in a private room wi th a visual divider between the partners so they could not see each other Participants were each photographed and then separated. Participants and their partners drew a picture of themselves and a picture of the other person in a randomly assigned order They were instructed to be as physically accurate as possible and to draw the person standing and facing the viewer. The participants then indicated with horizontal lines where they intended the shoulders, chest, waist, and hips to be so as to make potentially ambiguous drawings clearer. When the drawings were complete, the p articipants and their partners then indicated on th e Stunk ard Scale, a series of 9 figure drawings featuring men and women from very underweight to very overweight, which figure best represented their own bodies and the bodies of their partners. Participants and their partner s then completed the EDS 5, a 5 ques tion inventory which asks: Are you satisfied with your eating habits? Have you eaten to comfort yourself because you were unhappy? Have you felt guilty about eating? Have you felt that it was necessary for you to use a strict diet or other eatin g rituals to control your eating? Have you felt that you are too fat? Each question is rated on a 1 7 scale in which 1 equals 5. The minimum s core is 5 and the maximum score is
DRAWING TASKS FOR BODY PERCEPTION 19 35. A score of 5 indicates very low risk for eating disorders, and a score of 35 indicates very high risk. Participants also indicated their gender a nd sexual orientation as a free response to How do you define your gender? Some common examples include: and Finally, participants reported their height and weight before being debriefed. Body Image Perception Index To compare two drawings with photographs requires the images to be controlled for size. We therefore standardized the images by stretching or shrinking each image so that the height of the figure was 10cm. For example, a drawing with an 18cm tall figur e was shrunk 44.4% to set it to 10cm. Each of the horizontal measurements for this photograph was measured in the same way to create t he four Expected Measurements ( i.e S houlder, Chest, Waist, Hips) required to determine the Body Image Perception Index. The Body Image Perception Index (BPI) was found by dividing ([Drawn Measurement / Expected Measurement] x 100). This yielded the percent of the accurate proportion a p erson drew. A person who drew his shoulders 25% larger than their expected size would have a BPI of 125 for his shoulders. If he had drawn his shoulders 25% smaller than they really were, his shoulder s BPI would be 75. calculated, by subtracting 100 from each BMI and finding the absolute value of the result. This provides a positive score equivalent to
DRAWING TASKS FOR BODY PERCEPTION 20 the percent difference away from 100 for each measurement. In the afor ementioned case, wo would both be 25. Results Sexuality and Gender Of the 18 participants, 15 Hence, there was insufficient variability to compare statistics across gender and sexuality. Stunkard Scale, Body Mass Index (BMI), and Drawings Figure chosen on ( M = 3.96, SD = 1.11) was highly positively correlated with BMI ( M = 22.46, SD = 3.24) r = 0.85, p < 0 .0001 (Figure 4 ) Figure chosen was also positively correlated with ED S 5 score ( M = 16.67, SD = 7.29) r = 0.50, p = 0.035. FRS were more likely to have higher BMI and be at greater risk for eating disorders. BMI was not significantly correlated with Self Drawn Shoulders, Chest, or Waist, r (16) = 0.32, p = 0.20; r (16) = 0.37, p = 0.14; r (16) = 0.20, p = .42, respectively. H owever, BMI was positively correlated to Self Drawn Hips, r ( 16) = 0.53, p = 0.02. For with BMI. BMI was not correlated with Expected Shoulders ( M = 2.42, SD = 0.31), r (16) = 0.26, p = 0.30 ; with Expected Chest ( M = 1.83, SD = 0.16), r (16) = 0.73, p = 0.13 ; with Expected Hips ( M = 2.12, SD = 0.20), r ( 16) = 0.27, p = 0.28 BMI was correlated with Expected Waist ( M = 1.73, SD = 0.21), r (16) = 0.50, p = 0.04 For the mea sured dimensions in the photographs, only Waist was correlated with BMI.
DRAWING TASKS FOR BODY PERCEPTION 21 ( M = 2.42, SD = 0.31), r (16) = 0.23, p = 0.36 ; with Expected Chest ( M = 1.83, SD = 0.16), r (16) = 0.37, p = 0.13 ; with Expected Hips ( M = 2.12, SD = 0.20), r ( 16) = 0.22, p = 0.38 was correlated with Expected Waist ( M = 1.73, SD = 0.21), r (16) = 0.53, p = 0.02 Accuracy of Self Portraits Participants drew themselves more accurately for the lower half of their torsos than they did for the upper half. The proportions of the resized drawings participants completed of themselves were compared to the resized photographs of the participants. Four 2 tailed paired t tests were performed on the four measurements. Significant differences were found between Self Drawn Shoulders ( M = 2.12, SD = 0.63) and Expected Shoulders ( M = 2.40, SD = 0.31), t (16) = 2.81, p = 0.01; and between Self Drawn Chest ( M = 1.67, SD = 0.28) and Expected Chest ( M = 1.83, SD = 0.16) t (16) = 0.65, p = 0.03. No significant differences were found between Self Drawn Waist ( M = 1.59, SD = 0.32) and Expected Waist ( M = 1.73, SD = 0.21), t (16) = 1.48, p = 0.16. Differences between Self Drawn Hips ( M = 1.95, SD = 0.38) and Expected Hips ( M = 2.12, SD = 0.20) approached significance, t (16) = 1.81, p = 0.09. Body Perception Index Comparisons of measurements for one person drawn of him/herself and drawn of his/her partner yiel ded no differences. Four 2 tailed paired t tests were performed on the four calculated Body Perception Indexe s (BPI) : Shoulders, Chest, Waist, and Hips. Both drawings completed by the same person were paired. B oth drawings completed by each participant, one of him/herself and one of the paired participant, were included once in all analyses. No sig nificant differences were found: between Shoulder BPIs for Self ( M = 86.22, SD = 20.67) and Other ( M = 86.25, SD = 21.36) t (17) = 0. 01 p = 0. 99 ; between
DRAWING TASKS FOR BODY PERCEPTION 22 Chest BPIs for Self ( M = 91.81, SD = 15.03) and Other ( M = 94.24, SD = 15.53) t (17) = 0. 65 p = 0. 52 ; between Waist BPIs for Self ( M = 93.48, SD = 21.66) and Other ( M = 98.24, SD = 20.32) t (17) = 0. 81 p = 0. 43 ; between Hip BPIs for Self ( M = 92.33, SD = 18.75) and Other ( M = 91.53, SD = 17.89) t (17) = 0. 13 p = 0. 90 See Figure 2 for comparisons of the means At this level of accuracy participants were as accurate at drawing themselves as they were at drawing other people. Comparisons of Error measurements between drawings of Self and Other also yielded no differences. Four 2 tailed paired t tests were performed on the four Error calculations found, for Shoulders, Chest, Waist, and Hips. E ven with Error replacing BPI, no s ignificant differences were found between Self and Oth er for any of the four Errors: between Shoulder Error for Self ( M =18.59, SD = 16.20) and Other ( M =18.02, SD = 17.68) t (17) = 0.1 0 p = 0. 92 ; between Chest Error for Self ( M = 13.20, SD = 10.60) and Other ( M = 13.12, SD = 9.69) t (17) = 0. 02 p = 0. 98 ; between Waist Error for Self ( M = 16.52, SD = 15.01) and Other ( M = 14.01, SD = 14.41) t (17) = 0. 61 p = 0. 55 ; between Hip Error for Self ( M = 18.02, SD = 17.68) and Other ( M = 15.29, SD = 12.18) t (17) = 0. 50 p = 0. 62 (Figure 3 ) These measurements also suggest that participants were equally accurate at drawing themselves as they were others Eating Disorder Risk The EDS 5 was designed to work within a given community, so participants more than 1 SD above the mean in a representative sample are thought to be at risk for eating disorders. In this sample, the mean was 16.67 and the SD was 7.29. This means that a pe rson would need a score of over 23.96 to qualify. Only 3 participants had scores over 23.96. Their scores were 24, 25, and 27. This means that the population was at relatively low risk for eating disorders and should not be considered unhealthy in this regard.
DRAWING TASKS FOR BODY PERCEPTION 23 Score on EDS 5 was not significantly correlated with BPI of Shoulders for Self Drawings, r (16) = 0.03, p = 0.91; for Chest, r (16) = 0.06, p = 0.82; for Waist, r (16) = 0.39, p = 0.11; and for Hips, r (16) = 0.23, p = 0.36 (Figure 5 ) Score on ED S 5 was also not significantly correlated with Error of Shoulders for Self Drawings, r (16) = 0.04, p = 0.87; for Chest, r (16) = 0.29, p = 0.25; for Waist, r (16) = 0.38, p = 0.12; and for Hips, r (16) = 0.03, p = 0.90. The Paper Doll Hypothesis T o test whether p articipants are drawing the same figure twice the four Drawn Measurements by each participant were compared after being standardized Participant The differences in mean s were compared to 0 in both paired and unpaired t tests No significant differences were found between Drawn Shoulder Measurements for Self and Other, t ( 17 ) = 0. 28 p = 0. 79 ; between Drawn Chest Measurements for Self and Other, t ( 17 ) = 0. 65 p = 0. 53 ; be tween Drawn Waist Measurements for Self and Other, t ( 17 ) = 0. 84 p = 0. 41 ; between Drawn Hip Measurements for Self and Other, t ( 17 ) = 0.18 p = 0. 86 homogenous, this last test differences c ould be due to lack of real differences b etween partners The mean differences between grou ps of Expected Measurements were compared in both paired and unpaired t tests. No significant differences were found between Expected Shoulder Measurements for Self and Other, t (8) = 0.64, p = 0.54. No s ignificant differences were found between Expected Chest Measurements for Self and Other, t (8) = 0.45, p = 0.67. No significant differences were found between Expected Waist Measurements for Self and Other, t (8) = 0.70, p = 0.51. However, d ifferences
DRAWING TASKS FOR BODY PERCEPTION 24 between Expected Hip Measurements for Self and Other approached significance, t (8) = 2.19, p = 0.06. Disc ussion FRS, but for the photographed body dimensions, only with Waist size. Similarly, figure selected was significantly correlated with Waist size, but not the other dimensions. This average person than it does with actual body s hape. People who are aware that they have because they know that is where they fall relative to other weights. If so, it would be worthwhile to pursue this research with a sample of patients suffering from body image problems, i.e., Anorexia Nervosa, Bulimia Nervosa, or Body Dysmorphia. This population may need the freedom to express which aspects of their bodies make them most uncomfortable. Score on the Eating Disturbance Scale (EDS 5) was not correlated with accuracy. This finding could be because of the relatively healthy sample; only three participants were more than 1 SD from the mean. Typically, people without eating disorders are bett er at seeing themselves as they are ( Sand et al., 2011; Scagliusi, et al., 2006 ). Slade & Russell (1973) found that control populations did not display the same overestimation that patients with Anorexia Nervosa did. The healthy sample may also explain t he finding that p articipants were as accurate drawing themselves as they were drawing each other. Alternatively, it is also possible that the participants each have a standard mental image of a person that they draw whenever they are asked to draw someone Like a paper doll, this figure could be given different clothing, accessories, and facial features to resemble different people, despite
DRAWING TASKS FOR BODY PERCEPTION 25 being the same proportional figure. This is supported by the lack of significant differences between the adjusted m easurements between Self and Other drawings. However, t his lack could also be due to the homogen eit y in weights in the sample. According to the standards set out by the World Health Organization, only 3 participants in this study Therefore, it is impossible to tell whether participants were drawing a standard figure, each other, or themselves, because all three could have result ed in the same rate of errors. Partici pants in this study were accurate judges of their own body size relative to those of other people. This could be because of the relatively healthy sample; sampling people in a clinical environment would make this clear. If this task is an effective measu re of self perception, one would expect participants in a clinic suffering from eating disorders to draw themselves significantly less accurately than control participants (Channon, et al., 1990 ; Sand et al., 2011; Scagliusi, et al., 2006; Slade & Russell, 1973 ) We would expect the results from this study from any effective test of body perception in a healthy population ( Benninghoven, et al., 2007; Calden, et al., 1959; Sand et al., 2011; Slade & Russell, 1973 ) Further research should be conducted with different populations to determine whether this drawing task is an appropriate one to determine physical self perception.
DRAWING TASKS FOR BODY PERCEPTION 26 References American Psychiatric Association (2000) Diagnostic and Statistical Manual of Mental Disorders 4th edition, text revision. Washington, DC: American Psychiatric Association. Bennett, V. C. (1964). Does size of figure drawing reflect self concept?. Journal of Consulting Psychology 28(3), 285 286. doi:10.1037/h0042951 Benninghoven, D., Tetsch, N. N., Kunzendorf, S. S., & Jantschek, G. G. (2007). Perceptual body image of patients with Anorexia or Bulimia Nervosa and their fathers. Eating and Weight Disorders 12(1), 12 19. Retrieved from EBSCO host Brown, J. & Graham, D. (2008). Body satisfaction in gym active males: An exploration of sexuality, gender, and narcissism. Sex Roles, 59 (1 2) (2008). Brumback, R. A. (1977). Characteristics of the Inside of the Body test drawings performed by normal school children. Perceptual and Motor Skills 44(3, Pt 1), 703 708. Retriev ed from EBSCO host Buck, J. N. (1948). The H T P technique; a qualitative and quantitative scoring manual. Journal of Clinical Psychology 4317 396. doi:10.1002/1097 4679(194810)4:4<317::AID JCLP2270040402>3.0.CO;2 6 Calden, G., Lundy, R. M., & Schlafer, R. J. (1959). Sex differences in body concepts. Journal of Consulting Psychology 23(4), doi:10.1037/h0046643 Channon, S., de Silva, P., Hemsley, D., & Mukherjee, K. (1990). Body size perception and preferences in stable weight and improved weight anorexi c patients. International Journal of Eating Disorders 9(4), 403 408. doi:10.1002/1098 108X(199007)9:4<403::AID EAT2260090406>3.0.CO;2 2 Deaver, S. P. (2009). A normative study of children's drawings: Preliminary research findings. Art Therapy 26(1), 4 11 Retrieved from EBSCO host
DRAWING TASKS FOR BODY PERCEPTION 27 Ey, D. (2010). Body dissatisfaction and self esteem in males: Relationships with muscle dysmorphia. Dissertation Abstracts International 71, Retrieved from EBSCO host .. Fallon, A. E., & Rozin, P. (1985). Sex differences in perce ptions of desirable body shape. Journal of Abnormal Psychology 94(1), 102 105. doi:10.1037/0021 843X.94.1.102 Feyh, J. M., & Holmes, C. B. (1994). Use of the Draw A Person with conduct disordered children. Perceptual and Motor Skills 78(3, Pt 2), 1353 13 54. Retrieved from EBSCO host Garner, D.M., (1997). The 1997 body image survey results. Psychology Today, 30, 30 41. Goodenough, F. L. (1926). Measurement of intelligence by drawings. New York: Harcourt, Brace & World. Groth Marnat, G. (1998). Human fig ure drawings and house tree person drawings as indicators of self esteem: A quantitative approach. Journal of Clinical Psychology, 54 (2), 219 222. Honigmann, J. J., & Carrera, R. N. (1957). Cross cultural use of Machover's Figure Drawing Test. American Anthropologist 59650 654. doi:10.1525/aa.1957.59.4.02a00070 Koppitz, E. M. (1966). Emotional indicators on human figure drawings of shy and aggressive children. Journal of Clinical Psychology 22(4), 466 469. doi:10.1002/1097 4679(196610)22:4< 466::AID JCLP2270220430>3.0.CO;2 I Leon, G. R., Bemis, K. M., Meland, M., & Nussbaum, D. (1978). Aspects of body image perception in obese and normal weight youngsters. Journal of Abnormal Child Psychology: An official publication of the International So ciety for Research in Child and Adolescent Psychopathology, 6(3), 361 371. Doi:10.1007/BF00924739
DRAWING TASKS FOR BODY PERCEPTION 28 Levesque, M. J., & Vichesky, D. R. (2006). Raising the bar on the body beautiful: An analysis of the body image concerns of homosexual men. Body Image, 3(1), 45 55. doi:10.1016/j.bodyim.2005.10.007 Lo, W., Ho, S., Wong, B., Mak, K., & Lam, T. (2011). Validity and test retest reliability in assessing current body size with figure drawings in Chinese adolescents. International Journal Of Pediatric Obesity, 6 (2 2) e107 e113. doi:10.3109/17477166.2011.575146 Lynch, E., Liu, K., Wei, G., Spring, B., Kiefe, C., Greenland, P. (2008). The Relation Between Body Size Perception and Change in Body Mass Index Over 13 Years. American Journal of Epidemiology 169 (7), 857 86 6. doi: 10.1093/aje/kwn412 Matto, H. C. (2002). Investigating the validity of the Draw A Person: Screening procedure for emotional disturbance: A measurement validation study with high risk youth. Psychological Assessment 14(2), 221 225. doi:10.1037/1040 3522.214.171.124 Meyer, B. C., Brown, F., & Levine, A. (1955). Observations on the house tree person drawing test before and after surgery. Psychosomatic Medicine, 17 (6), 428 454. Milne, L. C., & Greenway, P. (1999). Color in children's drawings: The influe nce of age and gender. The Arts in Psychotherapy 26(4), 261 263. doi:10.1016/S0197 4556(98)00075 6 Peplau, L., Frederick, D. A., Yee, C., Maisel, N., Lever, J., & Ghavami, N. (2009). Body image satisfaction in heterosexual, gay, and lesbian adults. Archiv es Of Sexual Behavior 38 (5), 713 725. Rosenvinge, J. H., Perry, J. A., Bjrgum, L., Bergersen, T. D., Silvera, D. H. and Holte, A. (2001), A new instrument measuring disturbed eating patterns in community
DRAWING TASKS FOR BODY PERCEPTION 29 populations: development and initial validation of a five item scale (EDS 5). Eur. Eat. Disorders Rev., 9 : 123 132. doi: 10.1002/erv.371 Sand, L., Lask, B., Hie, K., & Stormark, K. (2011). Body size estimation in early adolescence: Factors associated with perceptual accuracy in a nonclinical sample. Body Image 8 (3), 275 281. doi:10.1016/j.bodyim.2011.03.004 Scagliusi, F., Alvarenga, M., Polacow, V., Cords, T., de Oliveira Queiroz, G., Coelho, D., & Lancha, A. r. (2006). Concurrent and discriminant validity of the Stunkard's figure rating scale adapted i nto Portuguese. Appetite, 47 (1), 77 82. doi:10.1016/j.appet.2006.02.010 Shenker et al. (1977) Self Perception in Obese Adolescents As Measured by Human Figure Drawings (Pediatric Research) Siconolfi, D., Halkitis, P. N., Allomong, T. W., & Burton, C. L. (2009). Body dissatisfaction and eating disorders in a sample of gay and bisexual men. International Journal Of Men's Health 8 (3), 254 264. doi:10.3149/jmh.0803.254 Slade, P. D., & Russell, G. F. (1973). Awareness of body dimensions in Anorexia Nervosa : Cross sectional and longitudinal studies. Psychological Medicine: A Journal of Research in Psychiatry and the Allied Sciences 3(2), 188 199. doi:10.1017/S0033291700048510 Stunkard, A. J., Sorensen, T., & Schulsinger, F. (1983) Use of the Danish adoption register for the study of obesity and thinness. Research Publications Association for Research in Nervous & Mental Disease 60, 115 120. Teglasi, H. (1980). Acceptance of the traditional female role and sex of the first person drawn on the Draw A Person T est. Perceptual and Motor Skills 51(1), 267 271. Retrieved from EBSCO host
DRAWING TASKS FOR BODY PERCEPTION 30 Tharinger, D. J., & Stark, K. D. (1990). A qualitative versus quantitative approach to evaluating the Draw A Person and Kinetic Family Drawing: A study of mood and anxiety disorder children. Psychological Assessment: A Journal of Consulting and Clinical Psychology 2(4), 36 5 375. doi:10.1037/1040 35126.96.36.1995 Thompson, J., & Altabe, M. N. (1991). Psychometric qualities of the Figure Rating Scale. International Journal Of Eating Disorders, 10 (5), 615 619. doi:10.1002/1098 108X(199109)10:5<615::AID EAT2260100514>3.0.CO;2 K Til lman, K.S. (2004) Human Figure Drawings: Evaluating Trends in Child Victims of Sexual Abuse
DRAWING TASKS FOR BODY PERCEPTION 31 A p p e n d i x Instruction: Please answer the following questions based on your experience in the last 30 days. 1. Are you satisfied with your eating habits? Very satisfied 1 2 3 4 5 6 7 Very unsatisfied 2. Have you eaten to comfort yourself because you were unhappy? Never 1 2 3 4 5 6 7 Every day 3. Have you felt guilty about eating? Never 1 2 3 4 5 6 7 Every day 4 Have you felt that it was necessary for you to use a strict diet or other eating rituals to control your eating? Never 1 2 3 4 5 6 7 Every day 5. Have you felt that you are too fat? Never 1 2 3 4 5 6 7 Every day Eating Disturbance Scale (EDS A new instrument measuring disturbed eating patterns in community populations: development and initial validation of a five item scale (EDS by Rosenvinge, J.H., Perry, J.A., Bjrgum, L., Bergersen, T.D ., Silvera, D.H., & Holte, A., 2001, Eur. Eat. Disorders Rev., 9 : 123 132. doi: 10.1002/erv.371
DRAWING TASKS FOR BODY PERCEPTION 32 Figure 1 Stunkard, A. J., So rens en, T., & Schulsinger, F., 1983, Research Publications Association for Research in Nervous & Mental Disease 60, 115 120.
DRAWING TASKS FOR BODY PERCEPTION 33 Figure 2 Average Body Perception Index for drawings of Self and Other. 0 10 20 30 40 50 60 70 80 90 100 Shoulders Chest Waist Hips Average BPI Average Body Perception Index Self Other
DRAWING TASKS FOR BODY PERCEPTION 34 Figure 3 Average Error for drawings of Self and Other. 0 5 10 15 20 25 30 Shoulders Chest Waist Hips Average Error Average Error Self Other
DRAWING TASKS FOR BODY PERCEPTION 35 Figure 4 1 2 3 4 5 6 7 8 9 15 20 25 30 35 Body Mass Index (BMI)
DRAWING TASKS FOR BODY PERCEPTION 36 Figure 5 Self Drawn Hip Dimensions compared to Body Mass Index. 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3 15 17 19 21 23 25 27 29 31 33 Drawn Hip Width BMI Dimensions of Drawn Self Hips compared to Body Mass Index
DRAWING TASKS FOR BODY PERCEPTION 37