Application of Mann-Whitney U test in educational research with SPSS

Page: 36-38

Sawinder Arora (G.N. College of Education for Women, Kapurthala, Punjab)

While analysis of data, the parametric statistical techniques, namely, t-test, analysis of variance, analysis of co-variance, multivariate analysis of variance, or regression analysis are preferred by the researchers due to the power they possess to reject the null hypothesis. These parametric inferential procedures are applied to make generalizations beyond the sample data. Although these statistical techniques rely on parametric assumptions, educational researchers give scant attention to the testing these assumptions. However, using parametric statistical techniques with disdain toward assumption testing can be hazardous. The sample statistics may not be a good estimation to the population parameters. An alternative to overcoming this situation is applying non-parametric techniques. Non-parametric tests are known as assumption-free tests. The present paper aims to discuss common non-parametric inferential procedures.