How to Measure Data Quality Accuracy

Data accuracy is one of the four components of data quality and sometimes gets the reputation of being one of the most difficult attributes to measure. It refers to whether the data values stored for an object are the correct values. If there is wrong data in the system, negative effects flows into reports. It’s the old mantra, “garbage in, garbage out.”

Two important requirements should be met for data to be accurate. First, it has to be the right value. Second, it has to represent the value in consistent form according to the data model and architecture designed.

There are several sources and causes of data inaccuracy. The most common of these comes during initial data entry from end users. If the user enters the record incorrectly, it’s going to be inaccurate. This is an element that can be overcome by having skilled and trained people doing the data entry. However, since mistakes can happen to everybody, data inaccuracy from data entry can be overcome by having components in the application detect typo errors, driving users to select dropdown list values or supplementing training materials with help text on the screen.

"Data Decay" can also lead to inaccurate data and happens when regular maintenance and data cleansing isn’t kept up. Many data values which are accurate can become inaccurate through time. For example the data found in people’s addresses, department and college assignments, ranks, work history and publication statuses can decay into inaccuracy. It’s important to build strong processes around these data points. When does it get updated and how often? Who has access to the right information? Is it possible to sync directly with a source system? Do faculty have the opportunity to identify/fix mistakes? Can a custom report be built to find conflicting data? 

"Data Movement" is another cause of inaccurate data. As data moves from one system to another, it could be altered to some degree. Values could change in transit or programs can intuitively remove leading zeroes or truncate large numbers.

Data Accuracy is a very important aspect in database management. While it’s difficult to catch everything, it’s important to have measures to minimize if not eliminate data inaccuracy. Measures to minimize can include overall reliance on reports from the system. Buy-in from faculty and staff, plus the knowledge that this system generates important reports on faculty data, can incentivize data accuracy across the board.

For more strategies to imperative your data accuracy, visit the article titled, Improving Your Data Quality | Accuracy.

Was this article helpful?
0 out of 0 found this helpful

Articles in this section

How to Contact Support
Click here to submit a ticket for the Support Team.
Watermark Academy
Click to access the Watermark Academy for free webinars, workshops, certifications, and free on-demand training