## Abstract The currently used criterion for sample size calculation in a reference interval study is not well stated and leads to imprecise control of the ratio in question. We propose a generalization of the criterion used to determine sufficient sample size in reference interval studies. The gen
Sample size requirements for multiple regression interval estimation
β Scribed by Douglas G. Bonett; Thomas A. Wright
- Publisher
- John Wiley and Sons
- Year
- 2010
- Tongue
- English
- Weight
- 353 KB
- Volume
- 32
- Category
- Article
- ISSN
- 0894-3796
- DOI
- 10.1002/job.717
No coin nor oath required. For personal study only.
β¦ Synopsis
Abstract
Sample size planning is one of the most important issues in the design of a study. Simple and accurate sample size formulas for a desired confidence interval width have been developed for many statistical procedures, but a simple and accurate sample size formula for the squared multiple correlation has been a notable exception. Several ruleβofβthumb sample size recommendations for a multiple regression analysis have been proposed over the years but none are satisfactory. Other approaches have focused on the construction of elaborate tables of sample size requirements, but these tables are both unwieldy and inadequate. We present a simple, accurate, and general method of approximating the sample size requirement for obtaining a squared multiple correlation confidence interval with desired precision. We also present a simple method for approximating the sample size needed to estimate unstandardized regression coefficients with desired precision. Copyright Β© 2010 John Wiley & Sons, Ltd.
π SIMILAR VOLUMES
Multiple comparisons are commonly made in epidemiologic and genetic research. How to appropriately adjust for multiple comparisons remains a controversial issue. This note demonstrates, however, that large increases in the number of comparisons has a limited e!ect on the sample size required to main
Methods for optimal sample size determination are developed using four popular multiple comparison procedures (Sche!e's, Bonferroni's, Tukey's and Dunnett's procedures), where random samples of the same size n are to be selected from k (\*2) normal populations with common variance , and where primar