Using Cluster Analysis for Data Mining in the Large-Scale Research

• In Event: Marriott Roundtable Session Seven
  In Roundtable Session: Data Analysis and Data-Driven Decision Making (Table 1)

Abstract

Clustering or grouping observations aims to differentiate observations by finding their similarities through a specific algorithm. This study aimed to use an unsupervised learning approach, clustering based on the covariates of the regression models, to deeper understand the main interest effect on the outcome variable. Applying two well-known clustering algorithms (K-means clustering and hierarchical clustering) with empirical examples, the findings suggested that using CA methods can be fruitful to elicit results, it was adequate to simplify or elaborate hypotheses on massive data, such as large-scale educational research, like the Program for International Student Assessment (PISA). The results were compared with that from the linear mixed models. Finally, the importance and the plan for later study was mentioned.

Author

• Xuran Wang, Michigan State University
  Presenting Author