91Ó°ÊÓ

In hypothesis testing, understanding Type I and Type II errors is crucial. These errors represent mistakes we can make when deciding whether to reject or accept a hypothesis.

A **Type I Error** occurs when we reject a true null hypothesis. In simpler terms, it's a false positive. We claim there is an effect when there isn't one. This error is controlled by setting a significance level, denoted by \( \alpha \), which is the probability of making a Type I error. In our example, this level is 0.01, meaning we are willing to accept a 1% chance of incorrectly rejecting the null hypothesis.

On the other hand, a **Type II Error** happens when we fail to reject a false null hypothesis. This is like a false negative, where we claim there is no effect when there truly is one. The probability of making a Type II error is denoted by \( \beta \). In problems like the example given, part of the task is to find \( \beta \) for certain parameters, such as when \( \mu = 94 \), where it was calculated to be approximately 0.4325.

Balancing the probabilities of these errors is essential in any test design. Lowering one type of error often increases the other, making it important to choose \( \alpha \) and \( \beta \) carefully based on the context of the test and desired outcomes.

The term "normal distribution" refers to the bell-shaped curve that describes how values of a variable are distributed. Many phenomena in natural and social sciences follow this pattern.

In our example problem, we assume that the distribution of the melting point is normal with a known standard deviation, \( \sigma = 1.20 \). What makes the normal distribution special is that it is completely characterized by its mean (\( \mu \)) and standard deviation (\( \sigma \)).

The properties of a normal distribution include:

• The curve is symmetric around the mean.
• Approximately 68% of the data lies within one standard deviation of the mean, about 95% within two, and 99.7% within three.
• The total area under the curve is 1, which represents the entirety of the event probability.

Recognizing that data follows a normal distribution allows us to employ z-scores in hypothesis testing, giving a standardized way to compare scores from different distributions by transforming them based on the mean and standard deviation.

This approach was key in the step-by-step solution. We calculated a test statistic \( z \) using the standard formula for normal distributions to determine whether to accept or reject the hypothesis.

Ensuring the correct sample size is crucial when planning experiments or studies because it affects the reliability and accuracy of results.

To determine the required sample size, we need to consider the desired levels of \( \alpha \) and \( \beta \). In the given problem, we wanted \( \beta = 0.1 \) with \( \alpha = 0.01 \). Here, calculating the correct sample size involves using the relationship between z-scores and the desired probabilities.

The usual formula to calculate sample size is based on making sure the test has enough power (which is 1-\( \beta \)) to detect a true effect. The exact formula involves:

• \( Z_{\alpha/2} \), which is the z-score that corresponds to half of the \( \alpha \) level, representing the cutoff for statistical significance in a two-tailed test.
• \( Z_{1-\beta} \), the z-score corresponding to the test's power, indicating the probability of correctly rejecting a false null hypothesis.

By rearranging the formula within the context of the specific test parameters, it was found that a sample size of 55 is needed to achieve the desired error probabilities for the given test of the melting points.