Snedecor and Cochran's Statistical Methods

by Koehler, Kenneth J.

• ISBN13: 9780813808642
• ISBN10: 0813808642
• Edition: 9th
• Format: Hardcover
• Copyright: 2016-01-01
• Publisher: BLACKWELL PUBLISHING

Summary

The much-anticipated ninth edition of this classic text/reference has been thoroughly updated. This new edition contains both the classic material you have counted on to drive home basic theory and points of reference; in addition, Dr. Koehler has added numerous new sections, figures and chapters to give students and professionals realistic data in which to learn from. Many previous chapters have been expanded. Just some of the new additions you will see in this new edition: Shapiro-Wilk and Kolmogorov-Smirnov tests for normality, expanded and more accurate approximation and discussion of sample size determination, and more accurate methods for constructing a confidence interval. Find new chapters on how to use logistic regression to model trends in proportions and how to apply Poisson regression models to count data. Another new chapter will introduce the exponential, log-normal and Weibull distributions, and discuss Kaplan-Meier estimators of survival curves and log-rank tests for comparing estimates of survival curves for several populations and much much more.

Table of Contents

Introduction

New section on stem-leaf displays, computation of sample percentiles, and box plots

Elimination of lower-case lettering to represent deviations from a sample mean

Expanded to indicate computer packages available to simulate normal distributions

Addition of Shapiro-Wilk and Kolmogorov-Smirnov tests; improved explanation of kurtosis and sample size determination

More accurate sample size determination for comparing two means

More accurate method for constructing a confidence interval when sample size is small

Improved explanation of rank sum test

New section on non-parametric methods for comparing several populations

Add formulas for 95% prediction intervals, added section on replication and lack-of-fit

No changes

Add section to provide references to use of log-linear models for higher dimensional tables

New chapter to show how to use logistic regression to model trends in proportions and how to apply Poisson regression models to count data

Add discussion of REML estimation

May only change chapter title

Expanded discussion of missing data and provide more recent references to the diagnostic techniques presented and discussions of the consequences of failures of model assumptions

Added section to explain the use of half normal plots to determine the significance of effects when there is no replication

Additional diagnostic procedures for assessing the fit of models and the influence of individual observations

Update references

Added section on the use of a model free procedure for fitting a smooth curve to a plot

New illustrations and updated references

No major changes

New chapter will illustrate the analysis of time to event data. It will introduce the exponential, log-normal and Weibull distributions, and the Kaplan-Meier estimators of survival curves and log-rank tests for comparing estimates of survival curves for several populations

Table of Contents provided by Publisher. All Rights Reserved.