Significance Test of Stress Value in Evaluating the MDS (Multidimensional Scaling) Growth Model Fit Using Permutation Test (Poster 49)

• In Event: Saturday Poster Session 1:15 pm
  In Poster Session: Longitudinal Studies SIG Poster Session

Abstract

The multidimensional scaling (MDS) model has been proposed as an alternative growth modeling method, called the MDS growth modeling. It has been used educational research for studying growth and change. Traditionally, the MDS model fit was assessed by stress value, which was used in descriptive way. In this paper, we evaluated permutation test as a method for the significance of the stress value (called p test) in assessing the MDS growth model fit. The results from simulation data and real-life example indicated that using permutation test in assessing the MDS growth model fit was viable and promising. Future studies can use the Mantel-test principle to test various structural hypotheses on the growth profile configuration in conjunction with p test.

Author

• Cody Ding, University of Missouri - St. Louis