What do the seemingly unrelated fields of professional sports and real estate have in common? Both make very effective use of data analytics, according to DSFederal Lead Data Analyst Echo Wang. Data analytics, as Echo explains it, includes a wide range of analysis techniques, especially analytics applications (i.e., Business Intelligence), such as data harmonization and data visualization. Data can illustrate cause-effect relationships, enable comparisons between programs or results, and even help to predict the future; and data analytics drives many government and business decisions. But how do decision-makers know that the data that they rely on is relevant, or even accurate?
As Echo puts it, “data tells the truth,” but only when it is valid. She explains, for example, that during warm weather, the number of bicycle-related injuries rises. This does not, however, prove that warm weather causes bicycle accidents. Proper data design will demonstrate that more people ride bicycles during warm weather, and the larger number of riders—not the higher temperatures—is the reason for the rise in the number of injuries. Data visualization—which visually illustrates data trends, using images or graphics—can provide a “snapshot” of a trend; or a detailed, nuanced picture. But that picture is meaningless, unless the data is up-to-date, accurate, and “clean.”
This is where data harmonization comes in. To harmonize data is to classify and name it, using consistent, agreed-upon, easy-to-understand labels. If data is properly harmonized, then database users will be able to find exactly the data they need, and avoid the data they don’t need. Most importantly, they will also know that their data is telling the truth. (Watch this DSFederal-produced video to learn more about data harmonization.)
Echo has led DSFederal efforts to harmonize data for several Federal customers, including the Department of Health and Human Services (HHS) and the Department of Transportation (DOT). For HHS, Echo helped to harmonize data housed in the Common Data Element Repository (CDER) Library, which hosts data from 26 grant-making Federal agencies. For DOT, Echo and her team standardized transportation safety data terminologies, established linkages and hierarchies among millions of data fields, and developed a data reporting prototype for DOT’s Bureau of Transportation Statistics (BTS), the principal source of transportation-related data for government and industry.
While transportation and grants analysis, among many other endeavors, rely heavily on data, perhaps no field is more closely associated with collection and analysis of data than public health. Public health researchers and analysts lead efforts to improve health outcomes and prevent the spread of disease. Through data collection, disease surveillance, and study of epidemiological events in a variety of populations, public health analysts uncover emerging health and disease trends.
It’s not enough, however, to demonstrate that an event has occurred or that a trend exists—public health analysts must also figure out why. Data analytics helps here, too. Using tools such as data dashboards, which provide real-time data visualization capabilities, data experts can pinpoint trends and quickly analyze them to determine possible causes. Data visualization can also facilitate improved healthcare by, for example, allowing hospital staff to accurately predict emergency room waiting times at different times of the day.
Data analytics, while useful, can also be expensive. A wide range of tools has emerged, which offer advanced analysis, visualization, and harmonization capabilities. These tools, such as Tableau and IBM Cognos, don’t come cheaply; and more importantly, technology can only do so much. Even the best and most expensive technology cannot substitute for human intellect and imagination. While a visualization tool can illustrate a data trend, only a person can understand why the trend is important, and how we should respond to it. DSFederal is lucky to have specialists like Echo, who help us to understand not just what, but why, and what next.