Testing on the Difference of Student’s Performance using Robust Methods
ANOVA is known to be adversely affected by non-normality and unbalanced design. Type I error and power rates are substantially affected when these problems occur simultaneously. Continuously using ANOVA under the influence of these problems eventually will result in unreliable findings. This study p...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IDOSI Publications
2012
|
| Subjects: | |
| Online Access: | https://repo.uum.edu.my/id/eprint/30955/1/WASJ%2017%202012%2066-71.pdf https://repo.uum.edu.my/id/eprint/30955/ https://www.idosi.org/wasj/wasj17(AM)12/12.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | ANOVA is known to be adversely affected by non-normality and unbalanced design. Type I error and power rates are substantially affected when these problems occur simultaneously. Continuously using ANOVA under the influence of these problems eventually will result in unreliable findings. This study proposed a robust procedure known as modified S1 and Ft methods. This procedure combines the S1 and Ft statistics with a popular robust scale estimator, MADn. A simulation study was conducted to compare the robustness (Type I error) of the method with respect to its counterpart from the parametric and non parametric aspects namely ANOVA and Kruskal Wallis respectively. Since the null distribution of S1 is intractable, bootstrap methods were used to give better approximation. The Ft used the approximation method. The findings were in favor of the S1 and Ft methods especially when the data were skewed. The performance of the methods was further demonstrated on real education data |
|---|