×

Warning

The RokSprocket Module needs the RokSprocket Component enabled.

Nonparametric Tests & Rank-Based Methods

Nonparametric Tests & Rank-Based Methods

Classical statistical methods such as ANOVA, linear regression rely on certain data distribution assumptions. Whenever they are not met, alternative methods such as nonparametric may be used. Learn about their principle, advantages and limitations.

Course Details

Learning Objectives

Learning Objectives:

Upon completion of this module, participants will be able to:

  • Understand the advantages & limitations of nonparametric tests
  • Assess when nonparametric tests should be used, based on the data and study objectives
  • Decide which nonparametric test to use given the data structure
  • Carry out the nonparametric data analysis
  • Interpret and report the results

Target Audience

Target Audience:

This module is intended for everyone involved in statistical data analysis (comparison of groups). It is especially intended for scientists who carry out experimentation with few replications.

Prerequisite

Prerequisite:

This module introduces the key ideas behind most commonly used nonparametric tests. It assumed that participants have no previous knowledge of statistics or that they have not used it for a long time.

Course Outline

Course Outline:
  • Conditions for the Application of Parametric Tests
  • Use of Nonparametric Tests
  • Differences between Parametric & Nonparametric Tests
  • Tests Adapted to the Various Experimental Designs
    • Case of a Single Group : Sign Test, Wilcoxon
    • Case of 2 Groups: Wilcoxon, Mann-Whitney
    • Case of More Than 2 Groups: Kruskal-Wallis
    • Case of Randomized Complete Block Designs: Friedman
    • Applications & Case Studies
    • Summary

Practical Info

Practical Info:

Recommended Duration: 0.5 to 1.0 day

Course Materials:

  • Course notes on statistical techniques
  • Sample datasets
  • Related Sessions

    • An easy intro to key statistical concepts. Learn about data visualization and outlier detection. Demystify statistical significance testing. An excellent introductory module and a solid basis for all other modules.