• Summary
• Related Documents
  • Active Bibliography
  • Co-citation
• Version History

Abstract:

Experimentation design is the main aspect that reflects and predefines the kind of analysis applied on derived data, in order for valid inferences to be determined from the experimental process. In classical experimentation, where all independent variables are in nominal or ordinal scale, various implementations of ANOVA are practiced and, as a result the significance of each internal data group is calculated. This well-known analysis procedure premises several assumptions concerning the residual normal distribution and equality of variances within each data group. On the other hand, resampling techniques powered by computer software provide an alternative approach free of any assumptions and theoretical necessities. In the present study, a software structure is presented designed to perform permutation tests on data composed out of one independent variable (factor), one response variable and one blocking factor. This is the typical data structure conclusion coming out of randomised block designs experiments. The software was designed as an Excel add-in that uses the flexibility of the spreadsheet environment and its calculation transparency. The software provides a parametric user interface and permits the user to perform large number of resampling trials according to various provided options. This software structure is presented in the present study along with a numerical experimental example used to illustrate its usage. 1. Introduction – About

Citations