Sequential Updating
What you'll learn: How to apply Bayes' Theorem repeatedly as new evidence arrives, refining your beliefs step-by-step.
The Core Idea
You've learned that Bayes' Theorem lets you update your belief about a hypothesis when you get one piece of evidence. But what happens when evidence arrives in stages—first one clue, then another, then another?
Sequential updating means applying Bayes' Theorem iteratively: your posterior probability after one update becomes the prior probability for the next update. Each new piece of evidence refines your belief further, building on what you learned from previous evidence.
How It Works
Think of a detective investigating a case:
- Initial belief (prior): Maybe there's a 20% chance the suspect is guilty based on background.
- First evidence (fingerprint found): Apply Bayes' Theorem → posterior becomes 60%.
- Second evidence (witness testimony): Use that 60% as your new prior → apply Bayes' Theorem again → posterior becomes 85%.
- Third evidence (alibi confirmed): Use 85% as the new prior → posterior drops to 30%.
Each update incorporates all previous evidence through the carried-forward posterior. You don't need to process all evidence at once—you can update as information trickles in.
Why It Matters
Sequential updating is powerful in real-world scenarios where:
- Medical test results come back one at a time
- Spam filters process email features sequentially
- Sensors gather data continuously over time
The order of evidence doesn't change the final result if you're updating correctly—whether you get Evidence A then B, or B then A, you'll reach the same conclusion.
Key Takeaway: Sequential updating chains Bayes' Theorem applications together, using each posterior as the next prior, allowing you to refine beliefs incrementally as evidence accumulates over time.