The descriptive results (means and standard deviations) related to the variables and their subscales are presented in
Table 1.
| Variable | Subscales | Males | Females | Total |
|---|
| Body image | Appearance evaluation | 20.77 ± 2.08 | 21.03 ± 2.29 | 20.88 ± 2.17 |
| Appearance orientation | 42.86 ± 2.71 | 43.76 ± 3.01 | 43.31 ± 2.87 |
| Fitness evaluation | 11.66 ± 1.62 | 12.13 ± 1.16 | 11.90 ± 1.42 |
| Fitness orientation | 42.26 ± 4.12 | 43.36 ± 2.78 | 42.81 ± 3.53 |
| Concern with weight gain | 3.70 ± 0.98 | 4.60 ± 1.45 | 4.15 ± 1.31 |
| Body area satisfaction | 26.86 ± 2.40 | 27.93 ± 3.65 | 27.40 ± 3.11 |
| Total body image | 1.46 ± 6.15 | 1.49 ± 8.73 | 1.48 ± 7.65 |
| Stress | Total stress | 1.19 ± 4.39 | 1.20 ± 8.86 | 1.20 ± 6.96 |
| Emotion regulation | Positive actions | 19.40 ± 1.79 | 19.40 ± 1.88 | 19.40 ± 1.82 |
| Controllability | 18.76 ± 2.20 | 18.13 ± 1.97 | 18.45 ± 2.10 |
| Expression of feelings and needs | 18.56 ± 1.99 | 17.40 ± 2.17 | 17.98 ± 2.15 |
| Assertiveness | 18.31 ± 2.36 | 17.67 ± 1.78 | 17.99 ± 2.100 |
| Well-being seeking | 20.67 ± 2.69 | 19.93 ± 2.19 | 20.30 ± 2.46 |
| Total emotion regulation | 95.48 ± 7.69 | 93.42 ± 4.89 | 94.45 ± 6.47 |
| Aggression | Physical aggression | 26.36 ± 2.45 | 25.93 ± 2.51 | 26.15 ± 2.47 |
| Verbal aggression | 14.73 ± 1.55 | 14.90 ± 1.34 | 14.81 ± 1.44 |
| Anger | 22.23 ± 2.66 | 21.96 ± 2.38 | 22.10 ± 2.50 |
| Hostility | 27.66 ± 1.74 | 26.26 ± 2.21 | 26.96 ± 2.09 |
| Total aggression | 90.40 ± 4.37 | 89.60 ± 5.47 | 90.00 ± 5.09 |
aValues are expressed as mean ± SD.
Is body image significantly related to stress, emotion regulation, and aggression?
To answer the abovementioned question, the Pearson correlation coefficient was used, the results of which are presented in
Table 1. This correlation test is applied whenever at least 2 variables are regarded and the variables are interval or relative. As indicated, body image was significantly and negatively correlated with stress (r = -0.259, P < 0.05). However, since the level of significance was greater than 0.05, there was no significant correlation between body image and emotion regulation (r = 0.156, P > 0.05). This is while there was a significant and negative correlation between body image and aggression (r = -0.258, P < 0.05).
Are the subscales of body image a predictor of stress among males and females with MS?
To answer the abovementioned question, stepwise regression analysis was applied. This type of regression analysis is used whenever examining the impact of several independent variables (interval or relative) on a dependent variable is targeted. The results demonstrated that in the first step, appearance evaluation was entered in the regression equation and was able to determine 0.14 of the variance in stress. In the second step, appearance evaluation and body area satisfaction were respectively entered in the regression equation and they determined a total of 0.21 of the variance in stress. In the third step, appearance evaluation, body area satisfaction, and fitness orientation were respectively entered in the regression equation and they determined a total of 0.27 of the variance in stress. In other words, the standardized beta coefficients indicated that one unit variation in the score of stress changed the variance in appearance evaluation by -0.39 in the first model, changed the variances of appearance evaluation and body area satisfaction by 0.31 in the second model, and changed the variances in appearance evaluation, body area satisfaction, and fitness orientation by -0.26 in the third model. For males with MS, in the first step, fitness orientation was able to predict 0.14 of the variance in the criterion variable (stress) and, according to the value of the adjusted R-squared, it had the greatest predictive power. Due to a lack of relationship with stress, the other subscales of body image were omitted. For the females with MS, in the first step, appearance evaluation was able to predict 0.18 of the variance in the criterion variable (stress) and, according to the value of the adjusted R-squared, it had the greatest predictive power. In the second step, appearance evaluation and body area satisfaction were able to predict a total of 0.29 of the variance in stress. Due to a lack of having any relationship with stress, the other subscales of body image were eliminated. Therefore, the obtained results revealed that none of the subscales of body image predicted emotion regulation.
Are the subscales of body image predictors of aggression among males and females with MS?
To answer the abovementioned question, the stepwise regression analysis was used. The results demonstrated that in the first step, fitness evaluation was entered in the regression equation and was able to determine 0.17 of the variance in aggression. In the second step, fitness evaluation and body area satisfaction were respectively entered in the regression equation and they determined a total of 0.23 of the variance in aggression. In other words, the standardized beta coefficients indicated that one unit variation in the score of aggression changed the variance in fitness evaluation by -0.43 in the first model and changed the variances in fitness evaluation and body area satisfaction by -0.26 in the second model. For the males with MS, in the first step, fitness evaluation was able to predict 0.18 of the variance in the criterion variable (aggression). In the second step, fitness evaluation and body area satisfaction were able to predict a total of 0.40 of the variance in aggression. Due to a lack of having relationships with aggression, the other subscales of body image were omitted. For the females with MS, in the first step, fitness evaluation was able to predict 0.17 of the variance in the criterion variable (aggression) and, according to the value of the adjusted R-squared, it had the greatest predictive power. Due to the lack of having any relationships with aggression, the other subscales of body image were eliminated.
To answer the following research question, i.e. are there any statistically significant differences between body image, emotion regulation, and aggression between males and females with MS? the obtained data was analyzed by the multivariate analysis of variance (MANOVA). This type of analysis is used whenever examining the impact of one independent variable on more than one dependent variable is targeted. To trust the results obtained from this type of analysis, the following conditions should be met. One of the assumptions of the multivariate analysis of variance is homogeneity of variance-covariance matrices. To assess this assumption, the box’s test of equality of covariance matrices was used (Box’s M = 11.08, F = 1.74, P = 0.107 > 0.05). The level of significance of the Box’s test was greater than 0.05; therefore, it could be concluded that the variance-covariance matrix was homogeneous.
To investigate the homogeneity of variances of the 2 groups, the Levene’s Test for equality of variances was applied. The results of the Levene’s test were not statistically significant for any of the variables under study (body image: F (1, 58) = -0.040, P = 0.84 > 0.05, emotion regulation: F (1, 58) = -0.002, P = 0.96 > 0.05, assurance: F (1, 58) = 5.93, P = 0.55 > 0.05, need to control thoughts: F (1, 58) = 2.74, P = 0.1 > 0.05, and cognitive self-awareness: F (1, 58) = 3.54, P = 0.06 > 0.05). Therefore, the assumption of homogeneity of variances was confirmed. This indicates that the following results are reliable.
The results of the MANOVA revealed statistically significant differences in the means of body image, emotion regulation, and aggression between males and females (F= 2.00, Wilks’ Lambda = 0.903, P = 0.05). Therefore, given the value of eta-squared (56.00), there were statistically significant differences between these 2 groups considering 3 variables of body image, emotion regulation, and aggression. Moreover, the power of the test (0.97) indicated adequacy of the sample size and its large effect size.
The results presented in
Table 2 showed the effects among the subjects (between the groups). Given the value of F, the differences in body image, emotion regulation, and aggression between the males and females with MS were significant at the level of P ≤ 0.05.
| Variable | Dependent | Sum of Squares | Df | Mean of Squares | F | Sig | Eta-Squared |
|---|
| Group | Body image | 101.40 | 1 | 101.40 | 1.75 | 0.005 | 0.029 |
| Emotion regulation | 62.83 | 1 | 62.83 | 1.51 | 0.005 | 0.025 |
| Aggression | 64.06 | 1 | 64.06 | 2.53 | 0.005 | 0.045 |
| Error | Body image | 3352.93 | 58 | 57.80 | - | - | - |
| Emotion regulation | 2410.99 | 58 | 41.56 | - | - | - |
| Aggression | 1465.93 | 58 | 25.27 | - | - | - |
Are there any statistically significant differences in stress between males and females with MS?
To answer the abovementioned question, the independent t-test was used. This test is used whenever 2 groups are needed to be compared with regards to an interval variable. With regards to stress, the results showed a statistically significant difference between the males and females with MS (t = -0.219, P < 0.05); therefore, given the mean scores of these 2 groups, it could be inferred that the females experienced higher levels of stress compared to the males. On other words, considering stress, there was a statistically significant difference between males and females with MS.