The relationship between sport participation, personality development, self-concept and self-esteem has repeatedly been discussed within the framework of legitimation discussions in the sector of school sport. The research shows contradictory findings (see [1
]). These contradictions concern, inter alia, the direction of causal associations of the investigated features. Sport and physical activity are often attributed in general with personality development impact, but this is not clearly and empirically established in this generalized point of view. Character development is often mentioned (e.g., in many physical education curriculum documents, such as [4
] see also [6
]) but no underlying concepts are explained to us that could enable empirical testing. The idea of “educational physical education” [7
] is, among other things, based on precisely these attributed effects. Thus, it remains an empirical question to prove or refuse these assumed effects, which cannot be answered in a general and simple point of view. Therefore, the aim of the present paper is to investigate the relationship between sports activity, motor performance and physical self-concepts. The extent to which the physical self-concept is influenced by stable personality traits is another subject of this study. The linking of the self-concept approach with the older, partly controversial trait approach should illuminate some new aspects. One further aim is the construction of a short scale to measure two important aspects of the physical self-concept, which are highlighted in this paper. One is the assessment of own physical attractiveness, the second is the assessment of own motor performance (general athleticism).
4.1. Scale Analyses
The body concept scale comprises a total of 15 items. The PCA resulted in an interpretable two-factor solution with 60% variance clarification that can be interpreted as a scale of physical attractiveness and a scale of general athleticism (Cronbach’s α each by 0.9). The overall finding was a good part-whole corrected item-total correlation (Table 2
). Taking into account the relevant pooling of the items, we can thus form two sum indices (sum scale ‘physical attractiveness’ and sum scale ‘general athleticism’, referred to hereinafter as ‘athleticism’). The totaled score on the ‘physical attractiveness’ scale varies from a minimum of 9 to a maximum of 45. The ‘athleticism’ scale varies from a minimum of 6 to a maximum of 30. The scale averages for ‘physical attractiveness’ were M
= 35.00 (SD
= 8.89, N
= 574) and for ‘athleticism’ they were M
= 24.31 (SD
= 6.88, N
4.2. Sex and Age
There are gender differences for both scales. The female respondents assessed themselves lower on average than the male respondents in both the trait ‘physical attractiveness’ and the trait ‘athleticism’. Accordingly, girls were on average less satisfied with their bodies and motor performance. As age-specific effects must often be supposed on psychometric scales, the variable age was also tested. The result of the analysis showed neither significant age-specific main effects nor corresponding significant interaction effects (Table 3
). Only the main effect gender showed a significant effect.
4.3. Sporting Activity
Children and adolescents who are active in sports assess themselves more highly both in the trait ‘physical attractiveness’ and in the trait ‘athleticism’ than those who are not active in sports (physical attractiveness: active in sports: M = 32.09, SD = 8.43, not active: M = 28.31, SD = 8.96; F(1, 981) = 20.65, p < 0.001, η2 = 0.02; athleticism: active in sports: M = 21.88, SD = 5.39, not active: M = 14.84, SD = 4.93; F(1, 982) = 181.97, p < 0.001, η2 = 0.16).
As significant effects are seen both in the case of the variable sex and in the variable sport participation, interactions between the two variables were tested using two-factor variance analysis with the two factors sex and sport participation (see Table 4
). There are two significant but rather weak main effects in the case of ‘physical attractiveness’ but no significant interaction effect, while for ‘athleticism’ all the effects were significant. While the main effect ‘sporting activity’ displayed a rather large effect size, the effect size of the main effect sex and the interaction effect are very small.
The two previously reported results were also tested with multivariate analysis of variance (MANOVA), which confirm the above (Sex: Pillai’s Trace = 0.02; F(2, 947) = 9.87, p < 0.001; η2 = 0.02; Sporting activity: Pillai’s Trace = 0.16; F(2, 947) = 88.04, p < 0.001; η2 = 0.02; Interaction Sex x sporting activity: Pillai’s Trace = 0.01; F(2, 947) = 3.12, p = 0.045; η2 = 0.01; all other main and interaction effects are not significant).
4.4. Motor Performance
The relationships between motor test performance and the two physical self-concept variables showed weak significant correlations for the assessment of physical attractiveness (r(395) = 0.31; p < 0.001) and moderate effects relating to the assessment of athleticism (r(395) = 0.41; p < 0.001). It was only for the individual item forward bend that no significant relationship was shown with the values on the scale for physical attractiveness.
Subsequently, the tests for relationships between the two sexes were carried out separately. This showed that the relationship is displayed somewhat more strongly for boys in relation to assessment of physical attractiveness (boys r(204) = 0.40; p < 0.001, girls (r(185) = 0.27; p < 0.001). With regard to athleticism, relationships tended to be greater for girls than for boys (boys r(205) = 0.53; p < 0.001, girls (r(185) = 0.58; p < 0.001), whereby the extent of the correlations did not differ significantly between the sexes in either case.
A multiple linear regression was calculated in order to test the level of influence of the motor performance and the mark in physical education. The results show that motor test performance has a substantially greater predictive value with regard to the self-assessment of athleticism for both sexes than the mark in physical education (Table 5
). The variance clarification was much weaker for the scale of physical attractiveness. Although the motor test performance had a significant, if weak, influence on the perception of own physical attractiveness for both sexes, the mark in physical education did not play a role for either.
4.5. Personality Traits
Multiple regressions were calculated in steps for both sexes to test the relationship between individual personality traits with the assessment of physical attractiveness and athleticism (Table 6
). The most noticeable thing here is the high negative importance of neuroticism. This shows the strongest relationship with the assessment of physical attractiveness and athleticism. A less strong relationship is shown in the case of extraversion with the assessment of athleticism. If we look at the relationships in a gender-specific process then there is a noticeable, slight gender-specific difference in the relationship with the scale of athleticism. For girls, the level of relationship between extraversion and athleticism is somewhat higher than that between neuroticism and athleticism, while this is reversed for boys.
The short scale for the measuring of the physical self-concept comprises a total of 15 items. The assessment of own physical attractiveness was measured using nine items, the assessment of own athleticism using six items. The item analysis resulted in a good scale property, so that the use of the measurement instrument for further analysis appears to be justified.
The assessments both of own physical attractiveness and own athleticism differ by gender. Girls see themselves somewhat more critically and/or they are somewhat less satisfied than boys [1
]. The findings also fit with possible gender-specific differences in motive for taking up sporting activities. Thus, we can certainly see higher participation in organised competitive sports for boys and somewhat higher participation in self-organised and non-competitive sport for the girls [47
]. However, the fact that no effects were shown with respect to age goes against the currently available findings [1
] and thus against the assumption of a clear differentiation of physical self-concepts during the puberty phase.
With regard to sport participation, there was a difference in both the assessment of own physical attractiveness and the assessment of own athleticism. Both were assessed somewhat more positively by adolescents who are active in sports than by those who are not active. It is, however, not possible to establish a causal relationship in the form of an effect of sports participation on the physical self-concept from the data. Competing explanations are, on the one hand, an influence on the physical self-concept through sporting activity (socialisation hypothesis) and, on the other hand, taking up sporting activity or sports participation because of a better physical self-concept (selection hypothesis). Burrmann [24
] was able to supply findings that were more in favour of the socialisation hypothesis. Overall, it was not possible to determine an important interaction between sport participation and sex.
The two recorded aspects of the self-concept also correlate with objectively measured motor test performance. Weak, but still statistically significant relationships are displayed between motor skill performance and the assessment of own physical attractiveness. As we have already argued that findings in the literature favour the socialisation hypothesis (self-concept changes through sports participation [24
]), the relationships between the objective motor performance level and measured self-concept aspects can be associated with the Skill Development model. This was also the conclusion of Asendorpf and Teubel [32
], who were able to prove this using longitudinal data from the LOGIK study [34
]. It was also shown in this context that motor testing with regard to self-concept had a higher predictive value for self-concept in the area of athleticism than the mark in physical education. This situation leads us to assume that children and adolescents assess their self-concept quite realistically and relate the assessment of their athleticism primarily to individual performance. The report of school performance seems to play a subordinate role in the formation of the physical self-concept. The rather insignificant influence of the mark in physical education in this context also corresponds to findings within the framework of the discussion regarding the diagnostic competence of teachers [48
]. If we pay attention to the specific aspects of self-concept that are studied in this context, then we see a tendency for boys to define their physical attractiveness more strongly based on their athleticism than girls. On the contrary, girls seem not to use the reference framework of school sports when assessing their physical attractiveness but rather to refer to other social contexts. Analogously to Rost and Sparfeldt [49
], we cannot, however, speak of general gender stereotypes here. The results rather point towards the necessity of a differential observation of sex in the context of the self-concept.
Overall, we were able to show that personality traits determine the physical self-concept to a certain extent. This aligns with the findings on the general self-concept [22
]. The influence of the trait of neuroticism is particularly highly substantial. The more strongly this trait is expressed (so the lower the emotional stability), the lower the assessment of own physical attractiveness and own athleticism. Reference group effects are of particular interest in further evaluation steps. Thus, it has already been possible to show that the sporting and/or physical self-concept of individuals in a high-performance environment (e.g., a school class with particularly athletic children) tends to be low [21
]. This corresponds to the ‘Big-Fish-Little-Pond-Effect’ (BFLPE) that was already discussed and proved in the 1980s with regard to the academic self-concept [20
]. Further research will examine supplementary influence variables e.g., parenting style and parenting goals.