Friday, March 24, 2017

OSP Information Team: Targeting Small Errors to Produce Better Big Data

Team members in the Office of Sponsored Programs are working hard and delving into what some might think of as tedious detail to ensure that users of their data can take its trustworthiness for granted.


Shenika Knox, Senior Data Analyst on OSP’s Information Team, knows a thing or two about data integrity.  She just spent months cleaning up customer tables from Oracle – thousands of records mis-classified and mis-coded customers dating back to Oracle’s adoption in 2001. 

In Sponsored Programs, these “customers” are their sponsors:  all of the entities who provide UVA with grants, such as foundations, government organizations, industries, etc.  Knox found that many customers had mistakenly been assigned to Sponsored Programs when they actually belonged to other departments. 

The upshot of having all of these extra customers in the system was that when the pre-award team set up new grants, they were faced with a massive, messy drop down list full of duplicates that took a long time to sort through and choose the correct sponsor.

“There were thousands of customers listed that belonged to other organizations, and that’s a problem, because it not only makes setting up PTAO’s difficult for us, but also, there are other University entities who are pulling this information about sponsors on research grants,” said Knox.

Knox, who has been a member of OSP for 9 years, took on the task of rectifying the situation.  With the help of teammate Nikita Mokhariwale, she set out to re-classify the records that didn’t belong, a process that involved a lot of detective work and close collaboration with UVA Enterprise Applications.

The cleanup that Knox led eliminates that time-consuming and frustrating part of the process for her colleagues, but the positive effects don’t stop there:  customer data can now be shared with confidence across the University.  ResearchNet, School of Medicine Clinical Research Improvement in Systems and Processes, and University Business Intelligence for research initiatives are just a few examples.  

Shenika Knox, Senior Data Analyst and data checking hero. 
Of her recent project to clean up thousands of mis-coded entries from Oracle, Knox said,
 “The whole time, I knew there was light at the end of the tunnel, but getting there was a challenge!
Vonda Durrer, Senior Director of Electronic Research Administration and lead on the ResearchUVA project, praised Knox and the Information Team for their work, noting that the more accurate customer tables also benefit other departments and units.

“From Student Financial Services and accounts receivable to the School of Medicine and the Office of the Vice President for Research, this is data lots of people want. We’re making sure the data they get is as accurate as possible,” Durrer said.

Next on the horizon for Knox and the Information Team is assisting with the new data visualization tool in ResearchUVA called JUICE.

“The work we do gives me insight into the business analytics that are behind the University’s effort to pull more research funds in. That’s something I’m proud to be a part of!”

No comments: