Statistical databases: Design of experiment structures

摘要

This paper discusses some special data patterns in relational databases, which are characteristics of the design and analysis of statistical experiments. First, we introduce the concept of order of functional independence in relational tables to provide a formal definition for relational data structure for statistical paired samples. We then introduce the statistical relational data patterns for randomized block designs and extended the definition of relational functional dependency so that, such data patterns can be extracted and analyzed using relational operations. The concepts of relational patterns are discussed. We have also provided some algorithms for selecting relational patterns from statistical relational databases.