Testing the differences of students’ scores between two groups
Normality and homogeneity are two assumptions that need to be fulfilled when dealing with t-test.However, these ideal conditions are rarely achieved in real world.Thus, this study proposed a robust technique that can deal with the non-normal data and variance heterogeneity.The p-value of T1 statist...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | en |
| Published: |
AENSI Publications
2012
|
| Subjects: | |
| Online Access: | https://repo.uum.edu.my/id/eprint/9385/1/4.pdf https://repo.uum.edu.my/id/eprint/9385/ http://www.aensiweb.com/jasr.html |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Normality and homogeneity are two assumptions that need to be fulfilled when dealing with t-test.However, these ideal conditions are rarely achieved in real world.Thus, this study proposed a robust technique that can deal with the non-normal data and variance heterogeneity.The p-value of T1
statistic that is combined with one of the popular robust scale estimators,MADn,Tn and LMSn was investigated.A simulation study was conducted to compare the robustness (Type I error) of the method with respect to its counterpart from the perspective of parametric and non parametric, i.e, the t-test and Mann-Whitney respectively. The performances of the method are demonstrated on real data on education.The findings show the favor of T1 method, especially for skewed data. |
|---|