HACETTEPE UNIVERSITESI EGITIM FAKULTESI DERGISI-HACETTEPE UNIVERSITY JOURNAL OF EDUCATION, sa.34, ss.168-177, 2008 (SSCI)
Experimental design is an important factor to assess the composite relations between dependent and interdependent variables in scientific research. To assess the associations precisely between the factors that involve in experimental design requires more sophisticated methods in the analysis process. CART is a non-parametric method that analysis complex relations between dependent and independent variables with group effects by classifying the sample into homogenous sub-groups and entering the model. Furthermore, CART has an advantage is that the analysis technique is free from the presumptions of multiple regression analysis such as normality, homogeneity, and interdependency of observations for parametric methods are ignored. The data was based on a survey among college students at Yuzuncu Yil University. The Limitedness Schemas Questionnaire, the General Self Efficacy Scale, and the Beck Depression Inventory were administered to 437 undergraduate collage students (313 males and 124 females). The data was analyzed by using CART analysis. Results were interpreted on a tree diagram obtained from statistical program with CART method. Relations between dependent and interdependent variables could be evaluated with the sub-group effects in the model. It was appeared that the sense of limitedness is a cognitive vulnerability factor for depression.