STATS 7 Lecture Notes - Lecture 14: Type I And Type Ii Errors, Null Hypothesis, Sample Size Determination
Document Summary
When null hypothesis is true, but it is rejected. When null hypothesis is false, but fail to reject it. The conditional probability of a correct decision, given alternative is true. The conditional probability of a type 2 error = 1 - power. Power = p(reject null | alternative is true) How to increase power (decrease type 2 error probability) Having more evidence makes it easier to show that the alternative hypothesis is true. The level of significance alpha is increased. It"s easier to reject the null when the cutoff value is larger. The actual difference between the true parameter and the null value increases. Easier to detect a true difference if it"s large. Must be taken into account when choosing alpha. If alpha is small it"s harder to reject the null. If alpha is large it"s easier to reject the null. If type 1 error is more serious, use smaller alpha.