91Ó°ÊÓ

The null hypothesis is a fundamental concept in hypothesis testing. It is a statement that assumes no effect or no difference in the context of a study. For example, if you are testing a new blood pressure medication, the null hypothesis might propose that the medication does not affect the blood pressure of patients. In the case of our exercise, it is expressed as follows:\[ H_0: \mu_{\text{before}} = \mu_{\text{after}} \]This means that the average blood pressure before taking the medication is assumed to be the same as after. The purpose of testing is to determine whether there is sufficient evidence to reject this statement.
If the null hypothesis cannot be rejected, it suggests that there is not enough evidence to say the medication has an effect. This is a crucial step because it sets a benchmark for comparison in hypothesis testing.

The alternative hypothesis offers a contrast to the null hypothesis. It represents the outcome that a researcher aims to support with evidence from the data. In the blood pressure medication example, the alternative hypothesis suggests that the medication does have an effect on blood pressure.
It is formulated as:\[ H_1: \mu_{\text{before}} eq \mu_{\text{after}} \]This means that the average blood pressure is assumed to change after medication, either increasing or decreasing. The role of the alternative hypothesis is to show the presence of an effect or a difference. Testing for the alternative hypothesis involves collecting data and determining if the observable results can reliably show a deviation from the null hypothesis.

• If the evidence is strong enough, the alternative hypothesis can lead to further research or potential changes in practice or understanding.

In any hypothesis test, the significance level is a predefined threshold that helps determine the likelihood of rejecting the null hypothesis given it is true. It is often denoted by \( \alpha \) and expressed as a percentage. In our context, the significance level is 1%, or 0.01.
The significance level plays a critical role as it sets the criterion for judging whether the observed data is inconsistent with the null hypothesis.

• For example, \( \alpha = 0.01 \) means that there is a 1% chance of rejecting the null hypothesis if it is true.
• This level is chosen based on the need for precision in the test. In a medical study, like a blood pressure medication test, a lower significance level (such as 1%) might be used to decrease the chance of a false positive, where it mistakenly shows that the medication is effective when it is not.

A two-tailed test is a method used in hypothesis testing when the alternative hypothesis involves a possibility of deviation in both directions from the null hypothesis. It accounts for both values higher and lower than the assumed value under the null hypothesis.In our example, the alternative hypothesis is \( H_1: \mu_{\text{before}} eq \mu_{\text{after}} \). This means we are interested in detecting both increases and decreases in blood pressure after medication.

• Unlike one-tailed tests, which only check for deviation in one direction, two-tailed tests are generally more conservative, requiring more extreme data to reject the null hypothesis.
• They are useful when the potential effects of a treatment or intervention could reasonably go in either direction, such as a drug that could potentially increase or decrease blood pressure.