A Complete Characterization of Triply Balanced Matrices with Applications to Survey Sampling.
Abstract
R x L triply balanced matrices arise in cross validation studies and in estimating the mean square errors of nonlinear statistics in many large scale survey samplings. It is shown that: (1) Any R x L triply balanced matrix and an orthogonal array OA(R,L,2,3; lambda) are one and the same object up to a possible notational change of the two symbols of the array to + and - respectively, (2) R is a multiple of 8 and L < or = R/2, and (3) The problem of the construction of R x L triply balanced matrices, 3 < or = L < or = R/2, is completely resolved modulo the existence of Hadamard matrices of order R/2. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1984
- Accession Number
- ADA147083
Entities
People
- A. S. Hedayat
- H. Pesotan
Organizations
- University of Illinois at Chicago