By using website you agree to our use of cookies as described in our cookie policy. Learn More

TDWI Research Perspectives on Data Literacy, Explainability, and Analytics Bias

December 1, 2021

Prerequisite: None

David Stodder

Senior Research Director


Data and analytics technology and cloud service investments are essential to achieving customer-centric goals, developing predictive insights, and running business processes efficiently. Yet, these investments could be for naught if organizations do not tend to human factors and governance issues that come with data collection, integration, and sharing as well as analytics developed with the data.

Data-driven decisions are playing an increasingly important role in organizations’ relationships with customers, consumers, healthcare patients, and citizens. Human factors involved in interpreting data and developing analytics affect these decisions. Governance rules and regulations are catching up, meaning that organizations need to be aware of relevant laws before they operationalize analytics and AI/ML. Social responsibilities, customer relationships, and business reputations are at stake.

This presentation will discuss TDWI research, industry trends, and regulatory concerns that tie together three key issues:

  • Data literacy - Data literacy addresses human aspects of how people interact with data. The primary goal is to raise individuals’ proficiency in understanding what data means and their ability to communicate and share analytical insights. A second goal is to increase responsibility and accountability for how users collect, integrate, prepare, and protect data. This is crucial to governance and regulatory adherence.
  • Explainability - Many organizations need to meet “explainability” requirements in data privacy regulations. In other words, they need to be able to explain how algorithms were applied to decisions such as determining a customer’s creditworthiness. This presentation will discuss how this regulatory principle affects data management and analytics.
  • Analytics bias - With predictive analytics affecting all kinds of decisions, organizations need to consider the problem of bias. Studies show that flaws in data collection, algorithms, and models could result in bias against minorities and other groups. Pressure is growing for decision makers to know whether they can trust analytics and AI/ML and whether organizations have accounted for and addressed potential bias problems.

Subscribe to Receive summit updates via email