A Statistical Approach to Vendor Selection
Abstract
A common problem that arises in practice is the comparison of several Bernoulli processes (or populations) with unknown parameters p sub l, .... p sub k, respectively, where the p sub i's denote the success probabilities. A particular realization of this problem is the critical issue of vendor selection. Deming (1982) notes the importance of vendor selection in a company's efforts to achieve high quality and productivity. In his 14 points, Demin's point 4 suggests the reduction of the number of suppliers to a subset of vendors who can furnish statistical evidence of dependable quality. Vendor selection involves a consideration of many aspects -- cost, service, reliability, and quality. Pettit (1984) described the approach that 3M Corporation uses in the evaluation of prospective suppliers. It consists of evaluating potential vendors in four areas: quality, price, performance, and facility capabilities. While quality is explicitly considered in this approach, it is not evaluated in a statistical sense. It is the intent of this article to indicate how statistics can be utilized as one objective evaluation tool in this decision setting.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1984
- Accession Number
- ADA149781
Entities
People
- G. C. Mcdonald
- Sumedha Gupta
Organizations
- Purdue University