Determining the Appropriate Sample Size for Confidence Limits for a Proportion
Abstract
Earlier papers by the present authors presented formulas for approximating with at least 0.999 relative accuracy the binomial confidence limits p sup bar and p sub bar based on a sample of size n with c 'defectives' drawn randomly from an infinite population with probability of p of a defective. The present article, in complementary fashion, presents substantially accurate procedures for determining appropriate sample size based on specifications as to the maximum anticipated value of c/n, the maximum desired value of the error margin which is the interval between c/n and p sup bar, and confidence level gamma.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 15, 1970
- Accession Number
- AD0714794
Entities
People
- Herman Burstein
- Theodore W. Anderson
Organizations
- Stanford University