The domain-specific risk attitude scale adopted for Turkish university students: Reliability and validity studies

The purpose of this research is to adapt the Domain-Specific Risk Attitudes Scale developed by Weber, Blais and Betz (2002) for Turkish university students. Besides the scores the students took from the scale is also compared according to gender.The research group comprises of students who study in the central campuses of Erciyes University, Gazi University and İnönü University in the spring term of the academic years of 2007 and 2008 and who volunteer for participating in the research. In total, 1048 students participated in the research and, of these students, 327 (%31) study in Erciyes University, 274 (%26) in İnönü University and 447 (%43) in Gazi University. Of the students who participated in the research (and stated their gender), 511 (%59) are female and 357 (%41) are male.In this research, in the analysis of the data, Comprehensive Exploratory Factor Analysis: CEFA-Statistica 8 package program is used. Relation of BARTÖ with social admiration is examined with Pearson Correlation technique. ?t? test is used for independent groups in the comparison of the points taken from the BARTÖ according to gender, students joining to research, members of climbing club and students taking the risk of climbing to the mountain.In this research, firstly, likert type with 40 themes and 6 dimensional the orijinal BARTÖ is translated to Turkish. However, 18 alternative themes are added to orijinal scale with considering cultural features of the Turkish university students. In order to determine construct validity of the Turkish BARTÖ, explanatory and confirmatory factor analysis is implemented to data. The results of analysis showed that the Turkish BARTÖ is 8 dimensional. According to the results of exploratory factor analysis made for bottom scales parcelling method, total variance, related to ?ethics? main factor, is explained by %35.4 with ethics in school, %15.3 with ethics in everyday life. Gambling games explains %39.9, investment explains %19.8 of the total variance related to finance sector. Security explains %32.1, health explains %14.5 of the total variance related to ?Health/Security? main factor. Entertainment main factor explains %32.4 and social main factor explains %25.4 of the total variance.Also, for the construct validity of the Turkish BARTÖ, Pearson Corelation technique is calculated in order to determine the relation of all the points taken from the BARTÖ by university students with the points taken from Social Admiration Scale (Kozan, 1983). This factor is founded as r=09, but it is meaningless on the level of 05. Besides, for distinguishing validity of the scale, the BARTÖ points of 34 climbing-club students and other 110 students are compared. The difference between two groups are founded meaningful (on the level of 001). Interior consistency and stability factors, related to all scale and sub-scales, are calculated in order to determine the reliability of the BARTÖ. Interior consistency factors calculated with Cronbach alpha tecnique are: for the whole of the scale: 89, for ethics sub-dimension in the school life: 68, for ethics sub-dimension in everyday life: 64, for sub-dimension of gambling games: 77, for investment sub-dimension: 75, for security sub-dimension: 65, for health sub-dimension: 50, for entertainment sub-dimension: 70, and for social sub-dimension: 57. Stability factor of the scale, which is calculated with again testing method, is founded as meaningful for the whole of the scale(r = .95). Those findings related to the validity and reliability of the BARTÖ shows that this tool is valid and reliable and can be used for the measurement of theTurkish students? behaviours about risk-taking.Also, in this research, points taken from the BARTÖ by university students are compared according to gender, because of the literature idea that there are differences between girls and boys about risk-taking behaviour. Points taken by boys from the BARTÖ are founded high in a significant level compared to girls, as it is in literature (t=9.83, p < .001). Findings of the BARTÖ about the comparison of points according to gender, supports former domestic and foreign literature research results.Key Words: University students, risk-taking, gender.