METHOD OF COMPLETE CONFOUNDING OF INTERACTIONS WITH SUB-BLOCKS IN FACTORIAL EXPERIMENTS WITH TWO, THREE, FOUR, AND FIVE LEVELS OF EACH FACTOR

Abstract

This paper presents the method of constructing and arranging factorial field experiments and other experiments with 2, 3, 4 or 5 levels of each factor according to the so-called rule of complete confounding. The definitions of the main effects and interactions, as well as the notion of generalized interaction, which is indispensable for understanding the method of confounding, are given in detail, making use of a consistent system of symbols, in four cases, the number of levels being 2, 3, 4, and 5 respectively. There is also described the general method of statistical interpretation of completely confounded experiments; as an illustration serves a detailed description of an experiment of the type 4 x 4 x 4 based on 4 sub-blocks. Stress is laid on the small number of calculations which have to be added when passing from the randomized blocks method to the method of complete confounding, as well as on the simplicity of the necessary modification of the analysis of variance.

Document Details

Document Type
Technical Report
Publication Date
Jan 07, 1963
Accession Number
AD0401075

Entities

People

  • Wiktor Oktaba

Organizations

  • Joint Publications Research Service

Tags

DTIC Thesaurus Topics

  • Analysis Of Variance
  • Combinatorial Analysis
  • Computing-Related Activities
  • Data Science
  • Information Science
  • Mathematical Analysis
  • Mathematics

Readers

  • Computational Linguistics
  • Computational Modeling and Simulation
  • Women's Health and Cancer Risk Research: African American Women and Pregnancy Outcomes.