Related Experiment Videos
A robust method for the analysis of experiments with ordered treatment levels.
J W McKean1, J D Naranjo, B E Huitema
1Department of Statistics, Western Michigan University, Kalamazoo 49008, USA.
Psychological Reports
|January 11, 2002
Summary
This study introduces a new statistical method for analyzing experiments with ordered treatments, offering a robust alternative to existing tests. The approach uses Spearman correlation and bootstrapping for reliable association measures and confidence intervals.
Area of Science:
- Statistics
- Experimental Design
Background:
- Analyzing experiments with ordered treatment levels requires specialized statistical methods.
- Existing parametric (e.g., Abelson-Tukey) and nonparametric (e.g., Terpstra-Jonckheere) tests have limitations for certain ordered alternatives.
Purpose of the Study:
- To present a robust statistical approach for analyzing experiments with ordered treatment levels.
- To offer an alternative to established methods like the Abelson-Tukey and Terpstra-Jonckheere tests.
Main Methods:
- The proposed method integrates Spearman rank-order correlation with bootstrap routines.
- This integration allows for the calculation of magnitude of association measures, p-values, and confidence intervals.
Main Results:
- The novel method provides a robust framework for analyzing ordered treatment levels.
- It offers advantages over five previously compared alternative approaches.
Conclusions:
- The Spearman correlation and bootstrap integration presents a powerful and versatile tool for analyzing ordered experimental data.
- This approach enhances the reliability and interpretability of results in experiments with ordered treatments.