Writing about Non-parametric Tests

Abstract

All of the statistical techniques described so far in this book have focused on parametric tests. If the data produced by a research project violate assumptions of parametric tests then researchers use non-parametric tests. The basic assumptions of all parametric tests are that the data are normally distributed, are interval or continuous, and the different groups have about the same amount of variance. Non-parametric tests are not as stringent in terms of assumptions as their parametric counterparts. For many statistical tests, there are non-parametric equivalents. This chapter examines some of the major non-parametric tests available to researchers and how they are commonly reported. The chapter covers the following topics:

• Technical information

  • Assumptions of non-parametric tests

  • Chi-squared

  • Mann-Whitney

  • Wilcoxon signed-ranks test

  • Kruskal-Wallis

  • Friedman’s ANOVA

  • Spearman’s rho

• Examples

  • Wilcoxon signed-ranks test

  • Mann-Whitney test

  • Kruskal-Wallis analysis of ranks

  • Friedman’s ANOVA