Adaptation of three questionnaires in Hungary

6.1.1 Adaptation of the Beliefs About Attractiveness Scale– Revised questionnaire in Hungary

The purpose of the first part of this study was to complete the psychometric evaluation of the Hungarian version of three questionnaires, one of which was the Beliefs About Attractiveness Scale – Revised questionnaire (BAA-R). The findings of the confirmatory factor analysis revealed that the theoretical two-factor structure of the questionnaire was not supported. Because of this, we carried out an exploratory factor analysis, which resulted in a new three-factor structure. The analysis showed that a three-factor structure with 14 items was a better fit in this case instead of the original two-factor structure with 19 items.

Besides the originally existing two factors (Importance of being attractive and thin and Importance of being physically fit and in-shape) a third factor was discovered of which factor the items were referring to the life fulfilment aspect of attractiveness. This factor was named the Life fulfilment aspect of attractiveness. The goodness of fit of the three-factor structure also proved to be acceptable, the internal consistency of the scales and the construct validity of the questionnaire was supported by the positive, significant relationships with the internalization subscales of the SATAQ-3 questionnaire. However, it is important to note that while these associations were moderate among women, regarding men they were only weak. We suggest a further psychometric examination of the BAA-R questionnaire on a wider range of population, especially among younger people, individuals suffering from EDs, or overweight people. The availability of this questionnaire in Hungarian contributes to the better understanding of societal values influencing beliefs regarding physical appearance and attractiveness.

6.1.2 Adaptation of the Physical Appearance Comparison Scale in Hungary

The second part of the first study aimed to adapt the Physical Appearance Comparison Scale (PACS) into Hungarian. After testing the single factor structure with CFA it was found that, because of the 4^th item, the questionnaire did not show a good fit. The reason for this is most likely that the 4^th item was the only reverse item in the questionnaire. Attention

was brought to the potential difficulties when reverse items are used and how these elements can alter the structure of a measure (Weijters, Baumgartner, & Schillewaert, 2013). After deleting the 4^th item and repeating the CFA, the model‘s goodness of fit greatly improved. The construct validity of the questionnaire was supported by the positive, significant relationships with the internalization subscales of the SATAQ-3 questionnaire and by the positive, significant relationship with the body dissatisfaction subscale of the EDI questionnaire. The observed associations were stronger among women (mostly strong and moderate associations), than among men. It could add value to the questionnaire if further psychometric examinations would focus on adolescents too, and on overweight, obese individuals and people who are suffering from eating disturbances. The availability of this questionnaire in Hungary might add to the understanding of body evaluation processes, and could deepen the understanding of the development of body image concerns.

6.1.3 Adaptation of the Social Comparison Scale in Hungary

The third goal of the first study was to adapt the Social Comparison Scale (SCS). In this case, the CFA did not support the single factor structure of the questionnaire. Instead, the exploratory factor analysis revealed a three-factor structure where the goodness of fit of the model was acceptable. The first factor (―Competence‖) was related to skills and abilities in association with social attributes based on which people carry out comparisons, higher scores on this scale meant that people rate themselves superior, more competent, talented, and stronger in comparison to others. The second factor (―Social acceptance‖) described the social acceptance and social belonging, higher scores here meant that the individual rated their social ranking compared to others as being someone who is more acceptable and more likely as an insider. The third factor (―Physical features‖) was related to the physical attributes of comparisons. Although the third factor was treated here as ―Physical features‖

it is important to note that one item (Undesirable/More desirable) had multiple meanings and could be explained as a social attribute too. In addition, because this factor had only two items, we did not use it in analyses in the current study. Internal consistency of the scales appeared acceptable. The construct validity of the questionnaire was supported by the positive, significant relationships with the RSE (self-esteem) scale. Associations

regarding both genders were strong. The availability of this questionnaire in Hungarian helps the understanding of the importance of social rank and evaluation processes regarding body image concerns. Future research could focus on carrying out further examinations including different age groups, or people suffering from existing mental health disorders such as mood disorders or depression.

As we saw, there were quite a few differences in the psychometric properties compared to the original version of the questionnaires. These differences can have multiple possible explanations. One reason could be the heterogeneity, size and the age range of the current sample. Originally, each validated questionnaire was first administered and then later used mainly in student groups, consisting of younger people of whom the majority were women.

For example, the original PACS was validated on 80 university female students (Thompson et al., 1991), the BAA-R questionnaire was first administered to 154 female undergraduates (Petrie et al., 1996) and when first validating the SCS, the sample consisted of 263 university undergraduates (Allan & Gilbert, 1995). Another reason for the discovered differences could be that the study was hosted online. Studies have documented that online data collection might modify in some cases the psychometric properties of questionnaires (Buchanan et al., 2005). They reported that the factor structure they found was different from the original paper-pencil version of the questionnaire. Reasons behind this could include the differences in the samples or the method of assessment. Furthermore, just like in our study, when Buchanan and colleagues (2005) assessed the psychometric differences between the online and offline versions of the same questionnaire, they found that the assessed sample was more diverse than the original one, which mainly included students.

The fact that the current study used different statistical packages (e.g. Mplus) to explore these psychometric properties of the questionnaires could also have an influence on the outcome and what factor structures and correlations were shown between variables. Lastly, the role of the culture might have influenced the outcome as well. Since all these questionnaires were validated on U.S. samples, the possible differences between the U.S.

and Hungarian culture and language could have contributed to the previously seen outcomes. Future research should focus on validating these measures offline as well as on samples that are more diverse. The above-validated questionnaires can be used in a wider

range of areas of psychology as well. The importance of social and appearance based comparisons, societal beliefs about attractiveness are important constructs considering social acceptance, and in general, when people evaluate themselves. Many areas of psychology could benefit from the existence of these measures. Human relationships, communication, and behavioural patterns are formed by underlying mechanisms that are many times influenced by perceptions from others and where people put themselves in comparison to peers.