91Ó°ÊÓ

When conducting a hypothesis test, the first step is to establish the **Null Hypothesis**. The null hypothesis, denoted as \(H_0\), is a statement that suggests there is no effect or no difference from a known standard or value. It serves as the baseline or default assumption that any observed effects are purely by chance. For instance, when evaluating societal beliefs, such as whether a certain percentage of adults believe rudeness is worsening, the null hypothesis provides a measurable parameter to test against.

• **Purpose**: The null hypothesis aims to retain the assumption that any experimental or survey result occurs simply by random chance and not due to actual effect or change.
• **Example Setup**: If a survey suggests that 75% of U.S. adults believe rudeness is worsening, then the null hypothesis \( (H_0: p = 0.75) \) reflects this scenario.
• **Verification**: The null hypothesis can never be "proven" true, but it can be either rejected or not rejected based on statistical evidence.

Thus, understanding the null hypothesis is crucial, as it frames the starting point of the hypothesis test and guides the decision-making process throughout the analysis.

The **Alternative Hypothesis**, denoted as \(H_a\), offers a statement contrary to the null hypothesis. It suggests there is an effect or a difference. This hypothesis is what the researcher typically wants to prove. In our exercise, the alternative hypothesis is \( (H_a: p > 0.75) \), suggesting more than 75% of U.S. adults believe rudeness is worsening.

• **Objective**: The alternative hypothesis challenges the null and attempts to show that the observed data come from a different distribution or have a pattern that \(H_0\) cannot explain.
• **Types**: There are typically two main forms:
  • **One-tailed test**: Investigates whether a parameter is greater than or less than the certain value stated by \(H_0\). Our example is a one-tailed test.
  • **Two-tailed test**: Looks for any significant difference, either greater or lesser, than the null hypothesis.
• **Analysis Implications**: Acceptance of \(H_a\) could imply new insights or discoveries, provided the evidence sufficiently supports it over \(H_0\).

Therefore, understanding the alternative hypothesis allows you to know what the research aims to demonstrate and sets the stage for testing procedures.

**Significance Level**, symbolized as \( \alpha \), is a critical concept in hypothesis testing. It represents the threshold at which you decide whether to reject the null hypothesis. A common significance level used is \( \alpha = 0.05 \), which translates to a 5% risk of wrongly rejecting \(H_0\). In our exercise, this is the level at which conclusions about Americans' beliefs on rudeness are drawn.

• **Risk Management**: Significance level addresses the likelihood of making a Type I error, which occurs when \(H_0\) is falsely rejected.
• **Importance**: The selection of \( \alpha \) depends on the research context, balancing between sensitivity (finding genuine effects) and specificity (not finding effects that aren't there).
• **Usage in Decision Making**: Once the P-value is calculated, compare it to \( \alpha \). If the P-value \( \leq \alpha \), then \(H_0\) is rejected in favor of \(H_a\). If not, \(H_0\) is not rejected.

Understanding significance level is essential because it dictates the stringency of the test and influences the confidence in the results. This measure ensures that researchers only accept findings that meet a predetermined level of statistical significance.