Relacije osobina ličnosti, situacionih varijabli i stilova odgovaranja na upitnike samoprocene

Abstract

This dissertation presents research aimed at investigating the status of response styles as latent variables that are stable across different questionnaires and can be encompassed by a general response style factor and at identification of dispositional and situational factors that can predict a participant's response style on self-rating questionnaires. The response styles considered in this research are extreme response style, acquiescence, disacquiescence, midpoint response style and socially desirable responding. We used a convenience sample of 541 adult citizens of the Republic of Serbia (59,7% female), between 18 and 86 years of age (M = 38.25 years, SD = 15.30). Participants from 53 settlements were included in the sample with most of them coming from three cities - Niš (32.3%), Užice (15%) and Belgrade (9.2%). As for the interrelations among the response styles, findings of our research are to a great extent in line with those found in the literature. Socially desirable responding was positively correlated with acquiescence and negatively correlated with the midpoint response style. Response style score stability across different questionnaires was investigated by Pearson's correlation coefficient for pairs of response styles estimates and the intraclass correlation coefficient (ICC) for general concordance of these estimates. The results supported stability of the response styles as measured by different means. We confirmed the existence of a general response style with the extreme response style, midpoint response style and socially desirable responding as its indicators and discussed further directions of its investigation. Differences in the response styles dependent on socio-demographic variables are in accord with those reported in the literature. Based on theoretical insights and empirical findings, a complex structural model of the relations between the epistemological constructs and HEXACO personality traits, on one side, and the response styles, on the other, was defined with the general response style modelled hierarchically. The results showed that the extreme response style, acquiescence, disacquiescence and midpoint response style may vary depending on the importance that a participant ascribes to a situation, as well as how items are formulated.