☠️ Error types
Types
accept H0 | reject H0 | |
---|---|---|
H0 = true | ✅ | 1️⃣ Type 1 error |
H0 = false | 2️⃣ Type 2 error | ✅ |
- Type 3: Wrong hypothesis, right answer
- cause: e.g. wrong Hypothesis Variables, poor theory
- Type 4: Right hypothesis, wrong answer
-
cause: e.g. wrong hypothesis test
-
Reducing errors:
- Increase 🎯 Accuracy, 🏹 Precision → less false positives (type 2 error)
- Increase 🛒 Recall / coverage → less false negatives (type 1 error)