Competing with High Quality Data: Concepts, Tools, and Techniques for Building a Successful Approach to Data Quality

Chapter 10 Information System Testing

10.0 Introduction

In previous chapters, we described methods to improve the quality of the data coming from various operational data sources and systems. If there are errors in these systems, then we might be working on the wrong data for evaluating DQ scores and initiating improvement activities. In order to verify that these systems are defect-free, we should perform measurement system analysis to ensure that the systems are highly reliable. Usually, the errors related to the systems can be fixed by studying the main effect of the factors or the combination effect of the factors interacting with these systems. The “system” here can be any software, analytical platform, or operational data source.

This chapter describes a methodology that can be used to test the performance of a given system and identify failing factors or signals that are responsible for poor information/data quality. The methodology described here uses the principles of robust engineering and orthogonal arrays to study two-factor interactions (combination effects) and main effects. Usually, it is sufficient to study two-factor combinations, because higher-order effects are small; hence, they can be neglected. This methodology aptly applies in the Improve phase of the DAIC approach because the main aim of this phase is to identify the failing factors and take suitable actions.

Generally, the technology teams or designers test the performance of a system by studying one ...