Developing an attitude scale towards negative integers for 7. and 8. grade students and analysis with logistic regression

The relationship between the attitude developed against mathematic course and the success in math course is highly focused research topic in literature. The success rate of a student varies proportionally with the positive manner developed to the subject.The negative integers topic is being taught at the seventh grade and the subject itself is found very hard to realize in real life by students. In this thesis, it is tried to achieved to develop an attitude scale against negative integers. The draft scale has been applied on 220 students selected among seventh and eighth grade, besides it has been verified by applying factor analysis. The cronbachs alpha reliability coefficient has been calculated for common reliability. The results has been evaluated on 95% confidence level. As a result, an attitude scale has been developed under one factor including 28 items.After the scale has been developed, the relationship between the data obtained at the first place and the success points at math has been observed, afterwards the points that students obtained from attiude scale has been divided into 4 equal intervals. These intervals has been managed as ordinal and the relationship of the student`s demographic data has been analyzed. Since the dependent variables were ordinal, the ordinal logistic regression technique has been preferred.A meaningful relationship has been observed on math grades taken between two consecutive terms which have been obtained by ordinal logistic regression models. The other model constitued with 7. class data is obtained that there is a meaningful relationship between mathematics attitude scale items and different categories of 24 items. It is received that it tend to show positive attitudes 7. class than 8. class and students taken support from outside of school that the other students against to negative integers in final model.Key words: Attitude scale , negative integers, ordinal logistic regression