Confidence Intervals — Visual Walkthrough
Confidence Intervals — Visual Walkthrough
For busy clinicians who prefer pictures over paragraphs. Move sliders, watch the story, learn the habit.
1) What is a CI?
95% CI idea: repeat the study many times → most intervals capture the true value.
2) Anchors: 0 (differences) vs 1 (ratios)
No-effect: 0
Rule: CI that crosses the anchor ⇒ inconclusive. CI fully on one side ⇒ effect present.
3) Make it precise
4) Forest plot & pooled diamond
Higher heterogeneity ⇒ wider pooled CI
5) SD vs CI
SD = patient-to-patient spreadCI = precision of the mean
