README.md

Confidence Interval (CI)

---

CI for z dist.

CI for t dist.

Sample vs Population

P-value on Normal Distribution

Confusion Matrix for Hypothesis Testing

Example for Confusion Matrix of AB Testing

z test score

t test score

z table (for alpha)

Usage of z table

t table (for alpha)

Usage of t table

Alpha (Type I Error), Beta (Type II Error) and Power

What is Power ?

• Probabilty of receting H0 when it is false
• 80% power means you have an 80% chance of getting a significant results when results when the effect is real
• Power tells you how likely you are to detect a real effect
• 80% power means you'll miss a real effect...
  • 20% of the time
• Want to find an initial effect and a replication given 80% power?
  • You will only find 64% of the time

Effect Size (ES) for Power Calculation

Effect Size (ES) Relationship with Power

Two-tailed, Left-tailed, Right Tailed Hypothesis Testing

A/B Testing

Two-sample z test for Difference Between Means

Example in Signal Detection with Diffrent Explanation

Hit, Miss, False Alarm, Correct Rejection and ROC Curve