PsychologicalReports, 1991, 68, 43-49.
O Psychological Reports 1991
AGE DIFFERENCES AND STRUCTURAL VALIDITY FOR T H E GREEK VERSION O F THE BODY CATHEXIS SCALE ' YIANNIS THEODORAKIS, GEORGE DOGANIS, KONSTANTINOS BAGIATIS Sport Psychology Laboratory Aristofle University of Thessaloniki Summary.-To examine age differences and the internal structure of the Greek version of the Body Cathexis Scale, 152 women 18 to 45 yr. of age and participating in physical fitness programs took the scale. High internal consistency was evident: in item analysis rs ranged from .29 to .68, split-half r was .89, and Cronbach alpha .92. Factor analysis yielded six factors: (1) Physical Fitness, (2) Total Appearance, (3) Health and Skills, (4) Face, ( 5 ) Ears and Chin, and (6) Extremities. These explained 52.5% of the total variance. The higher body satisfaction was observed on the factors, Health and Skills, Ears and Chin, and Extremities. Multivariate analysis of variance of body satisfaction across factors indicated no differences.
Body cathexis as part of self-concept concerns the relationships between the body and the self defined as "the degree of feelings of satisfaction or dissatisfaction with various parts or processes of the body (Secord & Jourard, 1953, p. 343). Studies showed that body satisfaction was significantly related to self-concept (Lerner, Karabenick, & Stuart, 1973; Rosen & Ross, 1968). A number of studies report significant relationships between attitudes toward the self and the body (Balogun, 1986; Berscheid, Walster, & Bohrnstedt, 1973; Darten, 1972; Tucker, 1983c; Zion, 1965). However, results of relevant studies are ambiguous in what ways variation in body satisfaction is related to age. Berscheid, et al. (1973) found that adolescents' appearance affects adult self-esteem, body image, and happiness and that people who were between 30 and 44 yr., but who had been beautiful adolescents, were the most happy. Among adolescents 14 to 18 yr. of age, satisfaction with the mouth, teeth, and lips increased significantly as age increased and a s i d a r but nonsignificant trend was found for peripheral body parts (Darnhorst, Littrell, & Littrell, 1988). Also, older respondents tended to have more positive body affect than younger ones (Gray, 1977) and postmenarcheal girls, as compared to premenarcheal, reported greater satisfaction with most characteristic female body parts on the Body Cathexis Scale (Koff, Rierdan, & Silverstone, 1978). Furthermore, preference for slim body shapes in women occurs at all ages (Furnham & Alibhai, 1983; Garner, Garfinkel, Schwartz, & Thompson, 1980). Other studies showed overestimation of body size decreased regularly 'Requests for reprints may be sent to G. Doganis, Ph.D., Sport Psycholo Laboratory, Department of Physical Education and Sports Science, Atistotie University oy~hessaloniki, Thessaloniki 54006, Greece.
44
Y. THEODORAKIS, ET AL.
with age (Kreitler & Kreitler, 1987) or an increase occurred in the evaluation of body parts from the third grade to the sixth grade (Vaughan, Stabler, & Clance, 1981). Body satisfaction declines during adolescence (e.g., hips, stomach, etc.) and is linked to psychological self-esteem and excessive dieting (Davies & Furnham, 1986). Powers and Erickson (1986) concluded that any body image may have been shaped a certain number of years previously, whether during adolescence or adulthood. On the other hand, women's attitudes about obesity were not affected by either age or levels of physical activity (Lai, 1984), and finally Fisher (1986) cited no variation in body cathexis related to age. In sports, evaluation of body cathexis is very interesting, since the better trained persons hold the more positive attitudes toward the self and the body (Balogun, 1987; Trujillo, 1983; Tucker, 1983a, 1 9 8 3 ~ ) Also, . athletes have been reported to score higher on the Body Cathexis Scale than nonathletes (Joesting & Clance, 1987; March &Jackson, 1986). For the estimation of body satisfaction, a number of body cathexis scales have been used (Berscheid, et al., 1973; Damhorst, et al., 1988; Mahoney & Finch, 1976; Secord & Jourard, 1953; Tucker, 1981), with sirnilar catalogs of body parts and functions. Other relative scales are the Verbal Checklist (Lerner & Korn, 1972) and the Body Image Satisfaction Scale (Rosen & Ross, 1968). Studies have shown that body cathexis is a multidimensional construct, as factor analyses of the scale have yielded four to six factors, explaining body satisfaction (Damhorst, et al., 1988; Mahoney & Finch, 1976; Tucker, 1981, 1983b). The validity of the scale has been extensively examined (Secord & Jourard, 1953; Berscheid, et al., 1973; Rohrbacher, 1973; Tucker, 1983~).In previous studies with Greek women (Theodorakis, Bagiatis, & Doganis, 1990) test-retest reliability was .90, Cronbach alpha .85, and a significant correlation ( r = .44, p < .001) was obtained with Rosenberg's Self-esteem Scale. The aim of this study was to examine the structure validity of the Greek version of the Body Cathexis Scale and to examine any changes occurring in body satisfaction among women of several ages. The null hypothesis was that satisfaction with the body does not vary with age.
Instrument In this study the Body Cathexis Scale (Secord & Jourard, 1953; Tucker, 1981) was administered. One each of 40 items responses were rated on a 5-point Likert scale "I feel very positive-I feel very negative." As other studies yielded sex differences (Berscheid, et al., 1973), this sample included only women (N= 152), 18 to 45 yr. of age (M= 30.4, SD = 9.6) who were participating in physical fitness programs. The sample was selected using
BODY CATHEXIS SCALE, GREEK VERSION
45
random stratified sampling, and all participants agreed to complete the inventory. The scale was readministered two weeks later. Data Analysis The SPSS/PC + Version 2 was used for the analyses. To examine the internal consistency of the scale, Cronbach alpha, split-half reliabilities, and item analysis were used. For the structural validity of the scale, factor analysis with orthogonal varimax rotation was performed. To determine whether the correlation matrix of the 40 items was appropriate for factor analysis two tests were used: Bartlett's test of sphericity, which showed that the variables = 3.11, p < ,000) and the Kaiser-Meyer-Olkin measure were independent of sampling adequacy, which gave a significant coefficient (KMO = .87) that is considered very satisfactory (Kaiser, 1974). Variables with factor loadings higher than .30 were retained. In the next step, the sum of factor scores were entered as separate variables into multivariate analysis of variance for age by body satisfaction to test for age differences. Multivariate statistics were preferred as they provide global evaluations of related body parts (Darnhorst, et al., 1988).
(x,,t
RESULTS Reliabilities Internal consistency by the split-half method was .89, Cronbach alpha was .92, and rs in the item analysis ranged from .29 to .68. All correlations were significant (p
O Psychological Reports 1991
AGE DIFFERENCES AND STRUCTURAL VALIDITY FOR T H E GREEK VERSION O F THE BODY CATHEXIS SCALE ' YIANNIS THEODORAKIS, GEORGE DOGANIS, KONSTANTINOS BAGIATIS Sport Psychology Laboratory Aristofle University of Thessaloniki Summary.-To examine age differences and the internal structure of the Greek version of the Body Cathexis Scale, 152 women 18 to 45 yr. of age and participating in physical fitness programs took the scale. High internal consistency was evident: in item analysis rs ranged from .29 to .68, split-half r was .89, and Cronbach alpha .92. Factor analysis yielded six factors: (1) Physical Fitness, (2) Total Appearance, (3) Health and Skills, (4) Face, ( 5 ) Ears and Chin, and (6) Extremities. These explained 52.5% of the total variance. The higher body satisfaction was observed on the factors, Health and Skills, Ears and Chin, and Extremities. Multivariate analysis of variance of body satisfaction across factors indicated no differences.
Body cathexis as part of self-concept concerns the relationships between the body and the self defined as "the degree of feelings of satisfaction or dissatisfaction with various parts or processes of the body (Secord & Jourard, 1953, p. 343). Studies showed that body satisfaction was significantly related to self-concept (Lerner, Karabenick, & Stuart, 1973; Rosen & Ross, 1968). A number of studies report significant relationships between attitudes toward the self and the body (Balogun, 1986; Berscheid, Walster, & Bohrnstedt, 1973; Darten, 1972; Tucker, 1983c; Zion, 1965). However, results of relevant studies are ambiguous in what ways variation in body satisfaction is related to age. Berscheid, et al. (1973) found that adolescents' appearance affects adult self-esteem, body image, and happiness and that people who were between 30 and 44 yr., but who had been beautiful adolescents, were the most happy. Among adolescents 14 to 18 yr. of age, satisfaction with the mouth, teeth, and lips increased significantly as age increased and a s i d a r but nonsignificant trend was found for peripheral body parts (Darnhorst, Littrell, & Littrell, 1988). Also, older respondents tended to have more positive body affect than younger ones (Gray, 1977) and postmenarcheal girls, as compared to premenarcheal, reported greater satisfaction with most characteristic female body parts on the Body Cathexis Scale (Koff, Rierdan, & Silverstone, 1978). Furthermore, preference for slim body shapes in women occurs at all ages (Furnham & Alibhai, 1983; Garner, Garfinkel, Schwartz, & Thompson, 1980). Other studies showed overestimation of body size decreased regularly 'Requests for reprints may be sent to G. Doganis, Ph.D., Sport Psycholo Laboratory, Department of Physical Education and Sports Science, Atistotie University oy~hessaloniki, Thessaloniki 54006, Greece.
44
Y. THEODORAKIS, ET AL.
with age (Kreitler & Kreitler, 1987) or an increase occurred in the evaluation of body parts from the third grade to the sixth grade (Vaughan, Stabler, & Clance, 1981). Body satisfaction declines during adolescence (e.g., hips, stomach, etc.) and is linked to psychological self-esteem and excessive dieting (Davies & Furnham, 1986). Powers and Erickson (1986) concluded that any body image may have been shaped a certain number of years previously, whether during adolescence or adulthood. On the other hand, women's attitudes about obesity were not affected by either age or levels of physical activity (Lai, 1984), and finally Fisher (1986) cited no variation in body cathexis related to age. In sports, evaluation of body cathexis is very interesting, since the better trained persons hold the more positive attitudes toward the self and the body (Balogun, 1987; Trujillo, 1983; Tucker, 1983a, 1 9 8 3 ~ ) Also, . athletes have been reported to score higher on the Body Cathexis Scale than nonathletes (Joesting & Clance, 1987; March &Jackson, 1986). For the estimation of body satisfaction, a number of body cathexis scales have been used (Berscheid, et al., 1973; Damhorst, et al., 1988; Mahoney & Finch, 1976; Secord & Jourard, 1953; Tucker, 1981), with sirnilar catalogs of body parts and functions. Other relative scales are the Verbal Checklist (Lerner & Korn, 1972) and the Body Image Satisfaction Scale (Rosen & Ross, 1968). Studies have shown that body cathexis is a multidimensional construct, as factor analyses of the scale have yielded four to six factors, explaining body satisfaction (Damhorst, et al., 1988; Mahoney & Finch, 1976; Tucker, 1981, 1983b). The validity of the scale has been extensively examined (Secord & Jourard, 1953; Berscheid, et al., 1973; Rohrbacher, 1973; Tucker, 1983~).In previous studies with Greek women (Theodorakis, Bagiatis, & Doganis, 1990) test-retest reliability was .90, Cronbach alpha .85, and a significant correlation ( r = .44, p < .001) was obtained with Rosenberg's Self-esteem Scale. The aim of this study was to examine the structure validity of the Greek version of the Body Cathexis Scale and to examine any changes occurring in body satisfaction among women of several ages. The null hypothesis was that satisfaction with the body does not vary with age.
Instrument In this study the Body Cathexis Scale (Secord & Jourard, 1953; Tucker, 1981) was administered. One each of 40 items responses were rated on a 5-point Likert scale "I feel very positive-I feel very negative." As other studies yielded sex differences (Berscheid, et al., 1973), this sample included only women (N= 152), 18 to 45 yr. of age (M= 30.4, SD = 9.6) who were participating in physical fitness programs. The sample was selected using
BODY CATHEXIS SCALE, GREEK VERSION
45
random stratified sampling, and all participants agreed to complete the inventory. The scale was readministered two weeks later. Data Analysis The SPSS/PC + Version 2 was used for the analyses. To examine the internal consistency of the scale, Cronbach alpha, split-half reliabilities, and item analysis were used. For the structural validity of the scale, factor analysis with orthogonal varimax rotation was performed. To determine whether the correlation matrix of the 40 items was appropriate for factor analysis two tests were used: Bartlett's test of sphericity, which showed that the variables = 3.11, p < ,000) and the Kaiser-Meyer-Olkin measure were independent of sampling adequacy, which gave a significant coefficient (KMO = .87) that is considered very satisfactory (Kaiser, 1974). Variables with factor loadings higher than .30 were retained. In the next step, the sum of factor scores were entered as separate variables into multivariate analysis of variance for age by body satisfaction to test for age differences. Multivariate statistics were preferred as they provide global evaluations of related body parts (Darnhorst, et al., 1988).
(x,,t
RESULTS Reliabilities Internal consistency by the split-half method was .89, Cronbach alpha was .92, and rs in the item analysis ranged from .29 to .68. All correlations were significant (p
