91Ó°ÊÓ

Hypothesis testing is a statistical method that helps us decide whether to reject a null hypothesis based on sample data. It starts with formulating two opposite statements: the null hypothesis (\( H_0 \)) and the alternative hypothesis (\( H_a \)).
The null hypothesis represents the status quo or a specific condition to be tested. For instance, it might claim that a population parameter, such as a proportion, equals a certain value. The aim is to determine if there is enough evidence in the sample data to reject this assumption.
The alternative hypothesis, on the other hand, is what you might want to prove. If the sample data provides enough evidence against \( H_0 \), you would reject the null hypothesis in favor of \( H_a \).
When conducting hypothesis testing, a threshold called the significance level is used, often denoted by \( \alpha \). Common choices for \( \alpha \) are 0.05 or 0.01, which determine how much risk you are willing to take for rejecting a true null hypothesis. This process usually involves calculating a test statistic based on your sample data, such as a Z-score in the case of Z-tests. This statistic helps to identify the P-value which will guide the decision to accept or reject \( H_0 \).

The P-value is a crucial element in hypothesis testing that helps determine the strength of the evidence in your sample data. It measures the probability of obtaining a test result as extreme, or more extreme, than the one observed, assuming the null hypothesis is true.
To calculate a P-value from a Z-test, you first need the Z-score, which is a standardized statistic indicating how many standard deviations your sample is away from the null hypothesis parameter value.
For a one-tailed test where \( H_a \) is greater than \( p \), the P-value is the area to the right of the Z-score. In contrast, if \( H_a \) is less than \( p \), the P-value is the area to the left. For two-tailed tests where \( p \) is not equal to a specific value, the P-value is twice the area of the smallest tail.

• For example, if \( z = 1.47 \) and \( H_a: p > 0.6 \), find the P-value using \( 1 - \text{norm.cdf}(1.47) \).
• If \( z = -2.70 \), and \( H_a: p < 0.6 \), the P-value is \( \text{norm.cdf}(-2.70) \).
• For a two-tailed test with \( z = -2.70 \), compute \( 2 \times \text{norm.cdf}(-2.70) \).

Finding the P-value helps you determine whether your sample data is unusual under the null hypothesis, guiding you in making the decision to reject or not reject \( H_0 \).

The normal distribution is a fundamental concept in statistics and vital for understanding hypothesis testing. Often depicted as a bell-shaped curve, it describes how data tends to be distributed around a central value.
In a normal distribution:

• The mean, median, and mode are all equal.
• The curve is symmetrical around the mean.
• Standard deviation dictates the width of the bell curve: larger values lead to a wider curve.

A standard normal distribution is a special case with a mean of 0 and a standard deviation of 1, often used in Z-tests. The Z-score, used in hypothesis testing, transforms the data into this standardized form, allowing easy calculation of probabilities (P-values) and comparison across various datasets.
Understanding normal distribution aids in assessing how typical or atypical your sample data is. It allows you to find areas under the curve (probabilities) related to Z-scores, which is crucial for determining P-values in hypothesis testing.

The alternative hypothesis (\( H_a \)) is a fundamental part of hypothesis testing. It is the statement you aim to support with your sample data. If sufficient evidence exists, you will reject the null hypothesis in favor of the alternative hypothesis.
The alternative hypothesis can be one of the following types depending on the research question:

• One-tailed: This considers only one direction of effect. For instance, stating \( H_a: p > 0.6 \) suggests that the population proportion is greater than a certain value.
• Two-tailed: This looks at deviations in both directions. \( H_a: p eq 0.6 \) implies that the population proportion might be either less than or greater than this value.

Choosing the correct alternative hypothesis is vital, as it defines how you will compute P-values and interpret your statistical outcomes. In hypothesis testing, the goal is to assess whether the observed data provides enough evidence to support \( H_a \), making it essential to clearly define the alternative hypothesis from the beginning of your analysis.