Design and Analysis of Experiments and More Realistic Techniques for Data Analysis.
Abstract
The research involved has been mainly focused on procedures -- on things to do with date -- on data analysis techniques. Some of the more innovative contributions include: graphical analysis of variance, configural polysampling, so far the only direct approach to robustness in actual finite-sized samples, compromise MLEs as a more flexible, often closely approximately, approach to robustness, calculations for many explicit randomizations as the trustworthy and stringent analysis for randomized experiment, simple devices for presenting motion using an overhead projector, work on pushback techniques and related ideas, effective confidence intervals based on only 2 of the observed values in a sample, new insight into multiple comparisons, fits with many exact-zero residuals sharing other good properties, deeper insight into, and robust procedures for, analysis of variance.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1986
- Accession Number
- ADA168921
Entities
People
- John W. Tukey
Organizations
- Princeton University