Data Quality Evaluation

Hensky consulting is active in the development of a new evaluation approach related to data quality.  The basic tenet of this approach is that many of the methodologies used by program evaluators can be applied to what might be referred to as data quality evaluations.  This idea may appear bold, as this field is commonly thought to be more the realm of data scientists and statisticians, it is argued that program evaluators can make a very real contribution in this area.

This idea will be discussed at the May 2015 Canadian Evaluation Meetings in Montreal.  The presentation will be based on this paper. This paper will be revised frequently as it is being currently read by several people before the meeting.

Our interest in data quality also extends to a deeper understanding of what the data means as we have found that frequently a lot of problems surrounding data arise from a superficial understanding of how it is collected and defined. Some evidence of our long-standing interest in the area is demonstrated in two papers from the 80s. The first paper illustrates the challenges of international data comparisons. The second demonstrates how difficult it is to translate a seemingly simple concept such as industrial output into a real measure.