91Ó°ÊÓ

Understanding the null and alternative hypotheses is the foundation of hypothesis testing. In any study, the null hypothesis (\( H_0 \) proposes that there is no effect or no difference, and it serves as the starting point for statistical testing. In the case of the ghost-believing US adults, the null hypothesis posits that only one-fourth (\( 0.25 \) of the population believes in ghosts. On the flip side, the alternative hypothesis (\( H_1 \) is the opposite of the null—it proposes that the true proportion is greater than the null hypothesis states; here, that more than 25% of the population are believers in ghosts. Through hypothesis testing, we essentially check the validity of the null hypothesis given the data.

To make the right decision between these contrasting hypotheses, proper execution of the test statistics is key, which will lead to a well-supported conclusion.

The test statistic is a standardized value that is calculated from sample data during a hypothesis test. It is used to make a decision about the null hypothesis by determining how far away the data are from what is expected under that hypothesis. Calculating the test statistic involves using a specific formula that takes into account the difference between the observed sample value and the expected value under the null hypothesis, after which it adjusts for the variability in the data and the size of the sample. The result is a z-score in our ghost-believing example. This score helps us understand where our sample data falls within the distribution that's assumed by the null hypothesis.

• For our scenario, a z-score of 2.98 means the observed proportion (31%) is 2.98 standard deviations above what the null hypothesis predicted, providing a numerical way to express the strength of our evidence.

A z-test is a type of statistical test that assumes the distribution of the test statistic under the null hypothesis approximates a normal distribution. It is especially useful for hypothesis tests concerning the population mean when the sample size is large, as in the case of our survey on belief in ghosts. The essence of the z-test is to assess whether there are significant differences from the hypothesized values.

Conducting a z-test typically involves the following steps: defining the hypotheses, selecting the significance level, calculating the z-score from the sample data, and then comparing the calculated z-score to a critical value from the z-distribution to make a decision about the null hypothesis.

In hypothesis testing, the significance level (\( \)alpha) sets the threshold for how extreme test results must be before we reject the null hypothesis. Commonly set at 0.05, or 5%, the significance level delineates the probability of rejecting the null hypothesis when it is actually true—this is known as a Type I error (a false positive).

In our example, with a chosen significance level of 0.05, we are stating that we are willing to accept a 5% chance of incorrectly rejecting the null hypothesis. If our test statistic exceeds the critical value related to our significance level, we reject the null and suggest that the alternative hypothesis may be true.

A one-sample proportion test is designed to compare the observed proportion from a single sample to a hypothesized population proportion. It's a specific application of a z-test when the variable of interest is a proportion. This is precisely what was done in our ghost survey when we tested whether the proportion of U.S. adults who believe in ghosts exceeds 25%.

For the test, we utilize the formula \(Z = (p - P0) / \sqrt{ P0(1 - P0) / n }\), with \(p\) being the observed sample proportion, \(P0\) the hypothesized population proportion, and \(n\) the sample size. In essence, the one-sample proportion test enables us to determine if the observed proportion (31% in our scenario) is statistically significantly different from our null hypothesis proportion of 25%.