91Ó°ÊÓ

The z-test for proportion is a statistical test used to determine if there is a significant difference between an observed sample proportion and a known population proportion. In this test, we compare our sample results to see if they are sufficiently different from our expectations, considering natural random variation.

To perform a z-test for proportion, we use the formula:

• \[ z = \frac{\hat{p} - p_0}{\sqrt{\frac{p_0(1-p_0)}{n}}} \]
• Where \( \hat{p} \) is the sample proportion,\ \( p_0 \) is the known population proportion, and \( n \) is the sample size.

This formula calculates a z-score, which tells us how many standard deviations away our sample proportion is from the population proportion. By comparing this z-score to a critical value, we can decide whether the observed deviation could be due to chance or if it’s statistically significant.

The null hypothesis, often denoted as \( H_0 \), is a statement that there is no effect or no difference, and it serves as the starting assumption for statistical tests. In many research scenarios, including the study of the effects of regular exercise on headache frequency, the null hypothesis asserts that the exercise program does not reduce the headaches related to diltiazem usage.

In our specific example, the null hypothesis is formulated as \( H_0: p = 0.12 \). This means that we initially assume the proportion of patients experiencing headaches remains at the prior established rate of \(12\%\), regardless of whether they exercise or not.

This hypothesis is essential as it provides a basis for testing whether there is sufficient evidence to believe a genuine change or effect is present. Failing to reject the null hypothesis means that any observed reduction in headaches is likely due to random chance rather than a result of the exercise.

The alternative hypothesis, noted as \( H_a \), presents a statement that contradicts the null hypothesis, suggesting that there is indeed an effect or a difference. In the context of hypothesis testing for our study, the alternative hypothesis proposes that regular exercise does contribute to a decrease in headaches for patients using diltiazem.

In mathematical terms, the alternative hypothesis for our case is written as \( H_a: p < 0.12 \). It indicates that the researchers expect the true proportion of patients experiencing headaches to be less than \(12\%\) if the exercise is effective.

Choosing the correct form of the alternative hypothesis is crucial:

• A left-tailed test, as in this example, checks for a decrease in proportion.
• It guides the direction and type of statistical test performed.

Accepting the alternative hypothesis usually implies rejecting the null hypothesis when enough evidence exists to support such a claim.

The level of significance, often denoted as \( \alpha \), quantifies the probability of rejecting the null hypothesis when it is actually true. Essentially, it represents the threshold for determining statistical significance. Researchers choose this level before conducting tests to minimize the risk of false positives, also known as Type I errors.

For our example with diltiazem and headaches, a \(1\%\) level of significance is used. This implies that there is only a \(1\%\) chance of mistakenly concluding that exercise reduces headache frequency when it actually does not.

The choice of significance level impacts the critical value of the statistical test. For instance, in a left-tailed z-test with \( \alpha = 0.01 \), the critical z-value is approximated as

• \(-2.33\).

Thus, any z-score lower than this threshold would lead us to reject the null hypothesis, offering stronger evidence against it.