91Ó°ÊÓ

Hypothesis testing is a statistical method used to make decisions about a population parameter based on sample data. This process typically involves two statements: the null hypothesis \( H_0 \) and the alternative hypothesis \( H_1 \). The null hypothesis represents a baseline statement, often one of status quo, while the alternative is a statement we seek evidence in favor of.

In the context of this exercise, we are testing whether the median impurity level is less than 2.5 ppm. Therefore, our null hypothesis is \( H_0: \widetilde{\mu} = 2.5 \), indicating that the median impurity level is equal to 2.5 ppm. Conversely, the alternative hypothesis \( H_1: \widetilde{\mu} < 2.5 \) suggests that the median impurity level is actually less than 2.5 ppm.

We then use a significance level \( \alpha \), usually 0.05, to decide on the strength of the evidence against the null hypothesis. If the \( P \)-value from the test is lower than \( \alpha \), we reject \( H_0 \) and accept \( H_1 \). This careful process ensures conclusions reflect genuine findings rather than random sample variations.

Normal approximation is a statistical method used to simplify complex distributions. It is particularly valuable in dealing with binomial distributions where parameters are large enough. In our case, it's applied to the sign test to approximate the binomial distribution of signs (\(+1\) or \(-1\)).

The normal approximation uses a normal distribution with mean \( \mu \) and variance \( \sigma^2 \). Specifically, for a test like ours with \( n = 18 \), the mean \( \mu = n/2 = 9 \) and the variance \( \sigma^2 = n/4 = 4.5 \). Through this approximation, we calculate a Z-score that allows us to assess the likelihood of observing our result (16 negative signs).

We compute the Z-score using \( Z = \frac{X - \mu}{\sigma} \), where \( X \) is our observed number of negative signs. This score is compared against the standard normal table to determine the \( P \)-value. A small \( P \)-value supports rejecting the null hypothesis, reinforcing the idea that the true median impurity level is below 2.5 ppm.

The median impurity level is a measure that represents the middle value of impurity content in a given sample. Unlike the mean, the median is less affected by extreme values, making it robust for skewed distributions.

In our problem, we are attempting to determine if the median impurity level is less than 2.5 parts per million (ppm). This measurement is crucial for ensuring quality and safety in chemical processes, as it indicates the presence of impurities within a sample.

By focusing on the median rather than the mean, the analysis minimizes the effect of any outliers and provides a more accurate representation of the underlying purity of the chemical product under scrutiny. This distinction is pivotal in industrial applications where precise impurity control is mandatory.