Overview: The session will introduce the three health data science tasks: description, prediction, and causal inference. Participants will begin by exploring descriptive methods, which involve summarizing and characterizing data to understand the burden and distribution of headache disorders within a population and discuss potential threats. Following this, the course will cover the fundamentals of prediction models, focusing on techniques used in predicting headache outcomes or treatment success. The third segment will address causal inference, explaining methods to determine cause-and-effect relationships in headache research. In the final part of the course, the concept of confounding will be introduced, highlighting how confounding can distort the observed associations between exposure and outcome. This session will equip attendees with a basic understanding of identifying and accounting for confounding in their research.