Solving Public Problems with Data

Data Analytical Thinking and Methods II: Correlation, Causation and Decision-Making

YOU WILL LEARN:

  • How to use data to make better decisions
  • The difference between association (or correlation) and causation
  • An introduction to the concepts of: conditional probability; independence; association; causal inference; and, statistical significance of results
Taught by
  • Daniel Goroff
38 mins

DESCRIPTION

This lecture discusses common inferential challenges and risks people face when working with data, including: thinking about conditional probability, the distinction between association (or correlation) and causation, what it means for events to be independent”, and how to claim a result is “statistically significant” is related to an implicit theory of randomness. The lecture closes by giving some practical advice about how to think through these concepts in your own domain.

RESOURCES

Causation vs Correlation by Sense About Science - An article that explores how conflating correlation with causation is one of the most common errors in health and science reporting.

Questions? Need help with a project?
Ask the public data science community at The Network of Innovators (NoI). NoI is a peer learning platform for finding practitioners with expertise and experience.

ASK AN EXPERT

LEARN NEXT

Barriers to Building a Data Practice in Government

Beth Blauer

Data Collaboratives: Creating Public Value by Exchanging Data

Stefaan Verhulst

Strengthening a Data Analytic Culture

Amen Ra Mashariki

Introduction to Evidence-Based Decision-Making: Examples in Economic Development

Quentin Palfrey

Data Analytical Thinking and Methods I: How To Define a Research Question and Introduction to Statistical Approaches to Draw Inference

Julia Lane

Machine Learning Applications for the Public Sector

Gideon Mann

Discovering and Collecting Data: Practical Advice for Government Managers

Carter Hewgley

Platforms and Where to Store the Data?

Arnaud Sahuguet