# Power and Beta Error

## 18 terms

α

1 - β

δ

### the significance level; power; the size of the difference in response that the researcher wishes to detect

Sample size is a function of:

### must increase to have a high probability (power) of finding the difference

As δ or α decrease, the sample size _____

### drop-out and drop-in rates

The sample size should be adjusted for _____

### α

Accounts for the chance that an observed difference in event rates will be large enough to lead the researcher to incorrectly reject the null hypothesis

### α

Used to establish the p value which reflects the probability of observing differences between treatment groups as large or larger than what would be reflected if the null hypothesis is true

### p ≤ α

The null hypothesis is rejected only if _____

### type II error

The probability of failing to correctly reject the null hypothesis

### β

The probability that the researcher accepts that he will miss a significant difference between treatment groups

### value of δ; sample size; α

β depends on several factors:

### power of the study

β is used to determine the _____

### power

The probability that a statistical test can detect a significant difference between populations when a difference truly exists

### power; a true difference of δ

Quantifies the ability of the study to find _____

### type I error

Incorrectly rejecting the null hypothesis

### consider the results of the study relevant for clinical decision-making; identify issues that merit caution in interpreting the results of the study

Power and β error are important to 1) determine whether to _____ and 2) to _____

### the detection of a lower difference than expected between the two groups

In the HC & FC trial, an increase in the sample size of nonresponders may have favored _____