EXPLORING ATTITUDES AND LEARNING STYLES OF STATISTICS LEARNERS USING PRINCIPAL COMPONENT ANALYSIS (PCA) AND MULTIDIMENSIONAL SCALING (MDS) APPROACH
Universiti Teknologi MARA (MALAYSIA)
About this paper:
Conference name: 13th International Conference on Education and New Learning Technologies
Dates: 5-6 July, 2021
Location: Online Conference
Abstract:
Learning and understanding statistics can enable a student to assess information and make proper decisions. Gauging students’ attitudes toward statistics and knowing their learning styles can help instructors to plan a better learning environment for them. This led to the objective of the study to identify the attitude dimensions and learning styles components that are more favourable to the students. A survey was conducted on a moderate number of students who enrolled in a Statistics for Business and Social Sciences course at a local public university in Malaysia. Survey of Attitudes toward Statistics (SATS) and Index of Learning Styles Questionnaires were used in measuring students’ attitudes toward statistics and their learning styles, respectively. Principal Component Analysis (PCA) was used to find an acceptable dimension of students’ attitudes toward statistics based on student’s characteristics. This has resulted in several factorization of items under respective dimensions. Felder-Silverman Model was used to evaluate students’ learning style preferences. This was further accomplished by using Multidimensional Scaling (MDS) technique to identify the similarities between SATS and Felder-Silverman components. From the MDS map, three different clusters that includes six attitude dimensions from the SATS and six Felder-Silverman learning styles components were formed. Domains in these three clusters were favourable to students and reflect their attitudes and learning styles in statistics. The outcomes would enable statistics educators to plan their teaching lessons tailor to the needs, attitude and learning styles of the students.Keywords:
Attitudes, Learning Styles, Learning Statistics, Principal Component Analysis (PCA), Multidimensional Scaling (MDS).