Likert scales are the most prevalent attitude scale. However, the fact that different items could be added up to produce same attitude values decreases the validity of the Likert scale results. Rough set data analysis could perhaps be conducted to overcome related limitations. Thus, this study seeks to investigate the employability of the rough set approach for the interpretation of Likert type attitude scales. Data was collected through a scale developed by using four sub-dimensions of the Fennema-Sherman mathematics attitude scale. In order to carry out the analysis, students were distributed to three groups of high, moderate and low in terms of their attitude to mathematics. Data was then converted to appropriate information tables for rough set analysis and lower and upper approximation sets were specified for these three groups. Accordingly, the potential membership of some students, who already belong to certain groups, to other groups was determined via rough sets. The mathematical value of to what extent each sub-dimension or any group can explain the total score was calculated. Through reduction of attributes, anxiety and usefulness sub-dimensions were found to be the indispensable sub-dimensions. The findings indicate that the rough set approximate can be used for the analysis of attitude scales.
Key words: Attitude scale, rough sets, data mining, approximations, data analysis.
Copyright © 2022 Author(s) retain the copyright of this article.
This article is published under the terms of the Creative Commons Attribution License 4.0