A Note on Selecting a Subset of Normal Populations with Unequal Sample Sizes.
Abstract
Let Pi sub 1, Pi sub 2,..., Pi sub k be k normal populations with common variance Sigma squared and let Pi sub i have mean theta sub i. Let n sub 1, n sub 2,..., n sub k given positive integers. Suppose n sub i observations are drawn from Pi sub i for each i = 1, 2,..., k, the authors require to select a nonempty subset of populations so that the probability that the subset selected includes a population associated with the largest mean is no less than P*(l/k < P* < L), a preassigned value. Two procedures of subset selection are given for the cases of known and unknown Sigma squared, respectively. Tables for the constants in the rule are given for some special choices of k, P* and alpha, the ratio of n sub i over the maximum n sub j. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1972
- Accession Number
- AD0748615
Entities
People
- Shanti Gupta
- Wen-tao Huang
Organizations
- Purdue University