Empirical Evaluation Methods for Pattern Languages: Sketches, Classification, and Network Analysis

Abstract

After elaborating design patterns in a certain field or domain building a pattern language will be the next step. This can be a challenging task, because a large amount of text with numerous cross-references and dependencies needs to become inherently consistent. In order to improve the quality of a pattern language it is necessary to analyse and evaluate all individual patterns as well as the pattern language as a whole. This paper provides an overview of existing evaluation methods for pattern languages and contributes further empirical methods that aim to assess and analyse pattern languages from different perspectives. Methods for sketching patterns, pattern classification, and the network analysis of pattern interrelation will be introduced and explained. The three methods are applied and tested in a case study about a pattern language for video learning environments.