Lecture 6A: Hypothesis Testing Flashcards
What is a statistical hypothesis?
A statement about the numerical value of an unknown parameter.
What is the null hypothesis (H0)?
Assume no difference/association.
What does H0: μ1- μ2=0 imply?
There is no significant difference
What does H0: ρ=0 signify?
There is no significant association.
What is the purpose of inferential statistics?
To determine whether the differences/associations found in a study are true or due to chance.
What are the five steps in hypothesis testing?
- State the statistical (null) hypothesis to be tested.
- Determine alpha value
- Run descriptive stat
- Run inferential stat and find p-value
- Make decision regarding null hypothesis
Null hypothesis assumes no diff/association
What is the alpha value in hypothesis testing?
The maximal acceptable risk of making a type I error, usually set at 0.05
Type I error: Incorrectly concluding null hypothesis is false
What is a type I error?
Incorrectly concluding that the null hypothesis is false when it is true.
What does p-value represent in hypothesis testing?
The probability of finding a result as extreme as the observed result, assuming H0 is true.
What decision is made if p < α?
Reject the null hypothesis.
What is the conclusion if p > α?
Accept the null hypothesis (No significant diff)
What is a type II error?
Accepting the null hypothesis when it is false, denoted by β
What is the relationship between sample size and power?
Larger sample size increases power (1-β).
What happens when the sample size is small?
Increased β leads to decreased power.
What is the conclusion if p = 0.036 in a study with α set at 0.05?
Reject the null hypothesis.
What is the risk of making a Type I error when p = 0.036?
3.6% chance of incorrectly rejecting H0.
What is the correct interpretation of a p-value of 0.02?
There is a 2% chance of making a Type I error if the null hypothesis is rejected.
What does the power of a study indicate?
The ability to find a difference when one truly exists.
What is the effect of increasing sample size from 63 to 126?
Increased likelihood of detecting a significant difference.
