91Ó°ÊÓ

A hypothesis test is a statistical method used to decide whether there is enough evidence to reject a null hypothesis. The null hypothesis, denoted as \( H_0 \), is a statement we initially assume to be true. For example, it might state that the population mean is a certain value. The alternative hypothesis, \( H_1 \), is what you want to prove.
In the exercise you viewed, part a had \( H_0: \mu = 23 \) and \( H_1: \mu eq 23 \), which means you're testing if the true population mean isn't 23. The process involves calculating a test statistic and comparing it to a critical value or, alternatively, calculating a p-value to make a decision.

• Null Hypothesis (\( H_0 \)): A statement of no effect or no difference.
• Alternative Hypothesis (\( H_1 \)): What we want to support, showing effect or a difference.
• Level of Significance (\( \alpha \)): The threshold for deciding whether to reject \( H_0 \), commonly set at 0.05.

By comparing our calculated p-value with \( \alpha \), we decide whether to reject the null hypothesis.

The test statistic is a standardized value used to determine the likelihood of observing the sample data if the null hypothesis is true. It's calculated using the formula: \[ Z = \frac{\bar{x} - \mu}{\sigma / \sqrt{n}} \] where \( \bar{x} \) is the sample mean, \( \mu \) is the population mean under \( H_0 \), \( \sigma \) is the population standard deviation, and \( n \) is the sample size.
In the original exercise:

• For part a, with \( \mu = 23 \), \( \bar{x} = 21.25 \), \( \sigma = 5 \), and \( n = 50 \), you'd use the formula to find the associated \( Z \) value.
• Each calculated \( Z \) value allows us to assess how far the sample mean is from the null hypothesis mean, in terms of standard deviations.

A large absolute \( Z \) value suggests our sample provides significant evidence against \( H_0 \).

A two-tailed test is used when the alternative hypothesis states that the parameter is not equal to a specific value. It checks for deviations in both directions—lower and higher than the hypothesized value. In hypothesis testing notation, it appears as \( H_1: \mu eq \mu_0 \).
Part a of the original exercise is a two-tailed test because \( H_1: \mu eq 23 \). This means we're concerned whether the mean is not equal to 23, on either side.

• The p-value here is calculated as \( 2 \times (1 - \text{normcdf}(|Z|)) \), doubling the tail probability due to checking both ends of the distribution.
• Such tests are less likely to detect an effect in a specific direction because they are spreading the significance level across both tails.

Two-tailed tests are typically used when any deviation from the null hypothesis is considered significant.

A left-tailed test is used when the alternative hypothesis specifies that the parameter is less than the null hypothesis value. It focuses only on the lower tail of the distribution. Notationally, it appears as \( H_1: \mu < \mu_0 \).
In the exercise, part b was a left-tailed test with \( H_1: \mu < 15 \). This implies we're only interested if the true mean is significantly less than 15.

• The p-value is computed using \( \text{normcdf}(Z) \), which directly provides the area under the curve to the left of \( Z \).
• Left-tailed tests are often used in quality control where a smaller value might indicate a problem, like lower-than-expected performance.

Remember, a significantly low p-value leads us to reject the null hypothesis only in this one direction.

A right-tailed test is appropriate when the alternative hypothesis suggests that the parameter is greater than the value proposed in the null hypothesis. It examines the upper tail of the normal distribution. In terms of notation, we write \( H_1: \mu > \mu_0 \).
Part c in the exercise fits this test type, with \( H_1: \mu > 38 \). It tests whether the true mean is significantly higher than 38.

• For this, the p-value is calculated as \( 1 - \text{normcdf}(Z) \), which captures the area in the upper tail.
• Right-tailed tests can be used in scenarios where exceeding a certain threshold is crucial, like improvements in test scores or production yields.

A significantly low p-value only solidifies our case against the null hypothesis in this direction.