What is Data Quality?
by Jason Cannizzo, APS TARC Staff
Over the past decade terms like big data, data analytics, and business intelligence have become commonplace, often heard in conversations between technology and business communities. The amount of data being collected has exploded in recent years, across all sectors of society. Coupled with our thirst for faster and better decision-making capabilities, an entire industry has been born. With so much attention focused on collecting data, and how to manage it, we sometimes neglect one of the most important ingredients - the quality of data. What happens when we collect the data, but don’t plan for quality? We’ve all heard the old adage, “Garbage in, garbage out,” which refers to the idea of inputting inaccurate, outdated, or inconsistent data that can lead to slow, costly, and misguided decisions.
What is Data Quality?
The term data quality refers to the characteristics associated with the data and the processes used to measure or improve the quality of data (DAMA International, 2017). High quality data means that the data fulfills its intended purpose by meeting the needs of its consumers. While there is no universal standard for data quality characteristics, most frameworks generally agree on common attributes. In this blog post we will focus on four common data quality characteristics that everyone should be aware of.
Completeness
Completeness refers to whether all required data is present. For data to be useful it must be complete. Think about something as simple as missing address information. This could result in staff inability to locate a client. At a more global level, missing data could result in incorrect conclusions especially when there is not enough data for an accurate representation. If your data system includes fields for assessment information, those fields should be completed.
Accuracy
Accuracy refers to the exactness and precision of data. It indicates the degree to which data is correct and represents reality. Accuracy can be difficult to measure. Most measures of accuracy rely on some type of comparisons to verify it is accurate. What would happen if a client’s medical diagnosis was not accurately documented? Failure could result in incorrect treatment, and inappropriate decisions being made. In other instances, inaccurate data could lead to misrepresentation of information.
Consistency
Consistency of data refers to the extent in which data is equivalent and consistently represented. Equivalency means the data is of equal value and has the same meaning. Imagine collecting data about full-time employees. Would the data be considered consistent and equivalent if some reported the number of positions filled, and others reported the number of positions allocated? The result would be inconsistent data, leading to a misrepresentation of information. Clear definitions, training, and standardized collection processes reduce inconsistencies with data.
Timeliness
Timeliness refers to the expectation that information is available and accessible to those who need it. Consider a scenario where case dispositions are not added in a timely manner. This could result in inadequately managed caseloads. Not every piece of data will have the same timeliness requirements, but every organization should document and clearly communicate expectation for timeliness of each specific piece of data.
Organizational Investment
Obtaining high quality data can be a strategic and momentous undertaking for any organization. Estimates differ, but some experts think that organization spend between 10-30% of revenue handling data quality issues (DAMA International, 2017). Organizations need to look at data as a strategic investment that goes beyond just a technology project. It’s an investment that should be integrated in the people, the culture, and the business itself.
It may seem like daunting challenge (and it is), but there are resources and tools available to help. Remember, improving the management and quality of your data is not likely to happening overnight. It requires intention, planning, coordination, and commitment. Start small and incrementally build the capabilities to manage and improve the quality of your organization’s data.
References
DAMA International. (2017). DAMA-DMBOK (2nd ed.). (S. Earley, D. Henderson, & L. Sebastian-Coleman, Eds.) Basking Ridge, New Jersey: Technics Publications.