Hypothesis Testing
A hypothesis statement is a true or false statement.
- When you are developing a hypothesis test, you are actually building a research question
- e.g., Did providing students with five extra study hours result in higher scores on the post test?
- When you convert your research question to a positive statement, you create an alternative hypothesis
- When you convert your question to a false or negative statement, you create a null hypothesis
Alternative hypothesis is the assumption that a relationship exists between two variables.
- e.g.
: There is a relationship between the five extra study hours given to students in classroom 105 and their higher scores. - the extra study hours made no difference at all, for either classroom
Null hypothesis is the assumption that a relationship does not exist between two variables.
- e.g.,
: There is no relationship between the five extra study hours and the higher scores in classroom 105. - the extra study hours made no difference at all, for either classroom
Errors
- we want to disprove the null hypothesis and prove the alternative hypothesis to be true
- but cannot reject it in error
- two types of errors
Type I error is an error that occurs when the correct hypothesis is rejected and the incorrect hypothesis is accepted.
- reject a null hypothesis that is actually true, creating a false positive
Type II error is an error that occurs when the incorrect hypothesis is accepted and the correct hypothesis is rejected.
- accept a null hypothesis that is actually false, creating a false negative