91Ó°ÊÓ

The significance level, often denoted by \( \alpha \), is a crucial element in hypothesis testing. It represents the probability of rejecting the null hypothesis when it is actually true. This is also known as the Type I error.

Choosing a significance level involves a compromise. A common convention is to use \(5\%\) or \(1\%\), indicating how willing we are to risk making this error.

• A \(5\%\) significance level means there is a \(5\%\) chance of concluding there is an effect, truly exists, when it doesn't.
• At a \(1\%\) level, we are more strict, being only \(1\%\) risk acceptable.

Significance levels help decide whether to reject \( H_0 \). It sets the threshold for what is considered unlikely if \( H_0 \) is true. The level must be set before performing the test to avoid biases.

In our example, decisions are made based on \(5\%\) and \(1\%\) levels, guiding us on whether the gathered evidence is strong enough for rejecting the null hypothesis.

A critical value is the threshold or boundary that separates the region where we would reject the null hypothesis from the region where we would not. It is determined based on the chosen significance level and the nature of the test (one-tailed or two-tailed).

• For a one-tailed test at \(5\%\) significance, we look for a critical value such that only \(5\%\) of the probability lies below it. This is approximately \(-1.645\).
• For a \(1\%\) level, it's \(-2.33\).

In a two-tailed test, we divide the significance level across both ends of the distribution. Hence at \(5\%\), we have critical values at approximately \(\pm 1.96\) and at \(1\%\), these values are \(\pm 2.576\).

To make our decision, we compare the test statistic with these critical values.

• If the test statistic is more extreme than the critical value, reject \( H_0 \).
• If not, we do not reject \( H_0 \).

Critical values thus form the foundation for hypothesis testing conclusions.

The null hypothesis, denoted as \( H_0 \), is a statement that there is no effect or no difference, and it is the hypothesis that researchers typically try to disprove. It often represents a situation of "no change" or "status quo."

When performing hypothesis testing, this is the assumption we initially consider to be true.
Usually, it posits that the population parameter contained remains at a specific value, like \( p = 0.65 \) in our exercise. Analyzing turns to whether evidence supports rejecting this hypothesis.

• Rejection leads to support for the alternative hypothesis.
• Failure to reject suggests insufficient evidence to support a change.

The null hypothesis serves as the baseline or reference for comparing the effect of interest.

In the context of our problem, it positions supportarily against random deviations being attributed to the assumed condition.

The alternative hypothesis, often denoted as \( H_a \), contrasts the null hypothesis and represents the outcome the researcher wishes to prove. It suggests that the population parameter differs from what is stated in \(H_0\).

In our exercise, there are two possible alternative hypotheses:
• \( p < 0.65 \) indicates that the proportion is less than assumed.
• \( p eq 0.65 \) means the proportion is different in some way.

The choice between a one-tailed and two-tailed test affects the form of \( H_a \).

• One-tailed tests suggest a specific direction of interest (e.g., less than or greater than).
• Two-tailed tests accommodate both directions, i.e., different without specifying if more or less.

Evaluating the alternative hypothesis involves examining the data to uncover support for \( H_a \) while considering chance variability.

A successful hypothesis test rejects \( H_0 \), hence lending credence to the alternative proposal.