How to build end-user data literacy skills, step by step

Top-notch hospital systems are filled with smart, knowledgeable people eager to use data. But first, they must improve their data literacy, as this case study shows.

Editor's note: The lack of data literacy skills among employees was called out at the recent Gartner Data & Analytics Summit as one of the four major obstacles preventing companies from thriving in a data-based economy. Here is how one data expert featured at the Gartner summit is facing this challenge.

The problem

Data, data everywhere -- and only a fraction of the workforce knows what to do with it.

That's what Jenifer Cartland, Ph.D., and her data and analytics team at the Ann and Robert H. Lurie Children's Hospital of Chicago discovered after launching a major initiative a few years ago aimed at providing self-service data analytics to the hospital's diverse workforce of healthcare givers, researchers and administrators. 

Jenifer Cartland, Ann & Robert H. Lurie Children's Hospital of ChicagoJenifer Cartland

"We really thought that if we just liberated the data, people would be able to use the data in an intelligent way and come away with the kinds of insights they were all clamoring for," said Cartland, vice president of data analytics and reporting at Lurie Children's and a research associate professor at Northwestern University's Feinberg School of Medicine.

The reality fell short.

"Every time we tried to give people data, something went wrong -- either they didn't know how to operate the system, or we had to do more training, or it was too simple for them," Cartland said at the Gartner event, recounting her group's multiyear effort to build data literacy skills.

The Lurie Children's Hospital workforce was filled with intelligent people eager -- and yet seemingly unequipped -- to capitalize on data, Cartland said. "We had to step back."

Cartland looked at research from the consultancy CEB, now part of Gartner, that showed organizations which derive insights from data have three characteristics:

  • Information that's accessible to users through features such as a centralized request system; interactive, shared dashboards; and easy-to-use tools;
  • Information that users want and find useful, such as agreed-upon key performance indictors and single-source-of-truth data supported by an expert analytical staff and a center of excellence for analytics; and
  • Capable employees.

"We kind of wondered if we were stuck" on No. 3, Cartland said. Another reason to re-examine assumptions? CEB found only 4.2% of healthcare workforces have the skills to be "high-performers in this kind of environment."


To gauge data literacy skills, Cartland surveyed the hospital's leadership tiers, a group of about 300 people. Only about half of respondents said they knew which metrics are important for gauging the hospital's success. "That was really shocking to me," Cartland said. The survey also revealed a "huge variability" in technical skills across the group and some surprising gaps: For "interpreting tables and graphs," data she assumed everyone would be comfortable with, only 75% of respondents said they possessed the necessary skills.

Based on the survey findings, Cartland's team then divided the population up into three "broad buckets of skills" that needed to be developed:

Independent analysts. These were analysts who typically were homegrown by their respective departments. An example would be the research assistant hired five years ago who was now expected to do high-level statistical analysis, Cartland said. Isolated, undertrained and lacking access to sophisticated data analysis tools, people in this group needed help with identifying "worthwhile data challenges," she said, as opposed to just "chasing the numbers of the month" for their leaders.

Leadership. Data literacy skills for this cohort centered mainly on interpreting tables and graphs. This group needed to understand and assess the data well enough to feel confident when using it to make decisions. They also needed be trained to ask the right questions about the data their analysts were presenting and be able to "cut through the hype" in vendor data.

Clinicians. This group, which included residents and fellows, needed a basic understanding of how to interpret research data, what kinds of data they needed for quality-improvement projects and how to translate the "big idea they saw at a conference" into something doable at Lurie Children's.


Cartland said one of the first steps in this multiyear journey to build data literacy skills was to establish "some uniformity" in data analysts' skills and job requirements. Many people in the hospital workforce had learned some analytics as part of their professional training or on a project, but the result was data silos and a lack of a common analytics language. "We wanted to create a multidisciplinary language around analytics and push people out of their boxes," Cartland said.

Establishing an Analytics Center for Excellence laid "claim to a place for analytics" in the hospital. Today, analysts hired by hospital departments typically co-report to Cartland's group, which also oversees job descriptions for new analyst hires.

Establishing a common data analytics language required having translators "on both sides," Cartland said. On one side are the "light quants," subject-matter experts in their disciplines -- nursing, sales or physical therapy -- who understand enough about analytics to talk to the professional analysts and translate back to their colleagues. On the other side are the professional analysts who have the communication skills to talk to nonanalysts. "This is not a skill all analysts have," Cartland said.

Cartland outlined other to efforts to develop data literacy skills across the healthcare organization:

  • Starting an internal blog, the Why Axis. "We came up with a bunch of things we wished people knew, and we just started writing blog posts on them."
  • Establishing office hours staffed by an analytics generalist. "[This is someone who] is midcareer, who's seen lots of different ways to look at numbers and knows our systems well enough to understand if people are using the right tools."
  • Contracting with two analytics coaches. "One works with the research and junior faculty, and one works with the quality improvement group."
  • Launching the "Analytics Network," a seminar series for Lurie Children's personnel and Northwestern University faculty. "It's not just another speaker series ... we have the speakers meet with groups we think might benefit from their experience."
  • Launching a short analytics course made up of five, two-hour sessions. "The first class will be nurses."
  • Offering in-person training.

Next steps include raising awareness of the blog and other analytics training, building out the analytics course syllabus and engaging other analytics trainers across the organization from data-oriented departments, such as finance.

Dig Deeper on IT applications, infrastructure and operations

Cloud Computing
Mobile Computing
Data Center
and ESG