Cluster analysis: a Tool for Analyzing Data in Quantitative and Mixed Method Studies

Subject Areas : Psychological Models and Methods

F. رادمهر ¹ , S.H علم الهدائی ²

1 - دانشجوی دکتری آموزش ریاضی دانشکده علوم ریاضی دانشگاه فردوسی مشهد.
2 - استاد آموزش ریاضی دانشکده علوم ریاضی دانشگاه فردوسی مشهد

Received: 2014-07-16 Accepted : 2014-07-16 Published : 2014-05-22

Keywords: Cluster analysis, Data mining, Psychology of Learning Mathematics,

Abstract :

In this study, cluster analysis as a descriptive data mining method for using in educational and psychological studies is presented. Grouping data in previous studies in these areas is done by median, mean, and range which has several weaknesses and limitations. Therefore, cluster analysis as an alternative method which has several applications in quantitative and mixed method studies is introduced. To achieve this, cluster analysis and steps for conducting cluster analysis (i.e., deciding on clustering variables, clustering procedure, selecting measure of similarity/dissimilarity, choosing clustering algorithm, deciding on the number of clusters, and validating and interpreting the cluster solution) are presented and illustrated by a sample data related to psychology of learning mathematics in order to familiarize researchers with this methods and also finding new results in this area. Obtained results indicate that cluster analysis can be used for investigating relationships between variables in quantitative research and finding potential candidates for interviewing in mixed method studies.