Business Intelligence and Data Quality

OLAP is short for "online analytical processing," and an OLAP cube, also according to Wikipedia, "is an array of data understood in terms of its 0 or more dimensions." [3] In this case, a business needs to design tests focused on certifying the relationships of parent-child data. It is the multi-dimensional capabilities that drive the power of the tool.

Figure 2: An OLAP cube

Conclusion
Database, warehouse, and BI testing are all much more challenging than testing the standard GUI application. This area of testing requires deep technical knowledge. Therefore, it is not uncommon for the ETL developers or DBA's to perform these tests.

The investment in BI tools and data analytics is deep, so ensure that investments in quality and testing are commensurate with the risks that the business is facing. Formalize methodologies including the spectrum of data warehousing and BI testing where necessary (measured by increased business risk). If a business is already dealing with regulated data or environments, then it is already aware of the more stringent controls that need to be in place.

Finally, the investment made in BI tools will not yield a good return on investment if the quality of the data is suspect. Directions and decisions made from poor data quality could be substantial and negatively impact your company.

References

1. "Extract, Transform, Load," Wikipedia, http://en.wikipedia.org/wiki/Extract,_transform,_load.
2. "Semantic Layer," Wikipedia, http://en.wikipedia.org/wiki/Semantic_layer.
3. "OLAP Cube," Wikipedia, http://en.wikipedia.org/wiki/OLAP_cube.