Question

The paper "I Smoke but I Am Not a Smoker" (Journal of American College Health [2010]: 117-125) describes a survey of 899 college students who were asked about their smoking behavior. Of the students surveyed, 268 classified themselves as nonsmokers, but said yes when asked later in the survey if they smoked. These students were classified as "phantom smokers," meaning that they did not view themselves as smokers even though they do smoke at times. The authors were interested in using these data to determine if there is convincing evidence that more than \(25 \%\) of college students fall into the phantom smoker category.

Short answer

The short answer depends upon the comparisons of the calculated test statistic value and the critical value. If the test statistic value is greater than the critical value, we reject null hypothesis, which means there is convincing evidence that more than 25% of college students are phantom smokers.

Step-by-step solution

Define hypotheses

The null hypothesis (H_0): \(p \leq 0.25\)\nThe alternative hypothesis (H_1): \(p > 0.25\) where \(p\) is the proportion of college students who are phantom smokers.

Calculate the sample proportion

The sample proportion \(\hat{p}\) is calculated as the number of successes (i.e., phantom smokers) divided by the sample size i.e., \(\hat{p} = \frac{268}{899}\)

Calculate the test statistic

The test statistic for hypothesis testing of a proportion is a z-score, which is calculated as follows:\nZ = \(\frac{(\hat{p} - p_0)}{\sqrt{\frac{p_0(1-p_0)}{n}}}\) \nwhere, \(\hat{p}\) is the sample proportion, \(p_0\) is the claimed proportion under the null hypothesis, and \(n\) is the sample size.

Determine the critical value and make a decision

Choose a significance level (\(\alpha\)), and find the critical value from the Z-distribution table. If the calculated Z-score is higher than the critical value, reject the null hypothesis.

Key concepts

Hypothesis Testing

In the realm of statistics, hypothesis testing plays a crucial role in enabling researchers to make inferences about population parameters based on sample data. It serves as a formal procedure for evaluating statistical evidence and helps them determine whether to accept or reject a statement known as the null hypothesis.

The null hypothesis, denoted as \(H_0\), often proposes no effect or no difference, and it specifies a value of a population parameter that the test aims to challenge. The alternative hypothesis, \(H_1\) or \(H_a\), opposes \(H_0\) and is what the researcher wants to prove.

For instance, in the scenario of phantom smokers, the null hypothesis posits that the proportion of phantom smokers in the college population \(p\) is less than or equal to 25%—formally written as \(H_0: p \leq 0.25\). The alternative hypothesis suggests that the true proportion is greater than 25%—\(H_1: p > 0.25\). The process of hypothesis testing involves collecting data and determining whether the results are consistent with \(H_0\) or if there is enough evidence to support \(H_1\).

To enhance clarity in the exercise solution, it's important to present the hypotheses in a manner that aligns with the research question and ensures the students understand the logical basis for accepting or rejecting \(H_0\) based on the sample evidence.

Sample Proportion

When examining a particular characteristic within a population, researchers often use a sample - a smaller, manageable group intended to represent the larger group. The sample proportion, symbolized as \(\hat{p}\), quantifies the fraction of the sample that displays the characteristic of interest.

In our example of phantom smokers, the sample proportion represents the fraction of surveyed students who smoke yet do not consider themselves smokers. It is calculated by dividing the number of students identified as phantom smokers (268) by the total number of students surveyed (899), resulting in a sample proportion \(\hat{p} = \frac{268}{899}\).

The accuracy of sample proportion as an estimate of the true population proportion depends largely on the sample size and how representative the sample is. Larger, random samples tend to provide more reliable estimates. Clearly understanding the concept of sample proportion is critical for students when interpreting the results of a survey and applying the findings to the broader population.

Test Statistic

Once the sample data has been collected and the sample proportion determined, the next step in hypothesis testing is to calculate the test statistic. This value helps in deciding whether the observed sample proportion is sufficiently unusual compared to the null hypothesis's claim.

For proportions, the test statistic used is usually a Z-score, which measures how many standard deviations an element is from the mean. Assuming the sample size is sufficiently large and the conditions for the Central Limit Theorem are met, the test statistic for the proportion is computed using the formula: \[ Z = \frac{(\hat{p} - p_0)}{\sqrt{\frac{p_0(1-p_0)}{n}}} \] where \(\hat{p}\) is the sample proportion, \(p_0\) is the proportion under the null hypothesis, and \(n\) is the sample size.

In the scenario of the phantom smokers, the test statistic would tell us how far the sample proportion \(\hat{p}\) is from the null hypothesis value of 0.25. This calculation gives an objective measure to compare against a critical value based on the chosen significance level, assisting in the decision to reject or not reject \(H_0\). Explaining the test statistic in simpler terms and illustrating how it fits into the hypothesis testing process can greatly assist learners in navigating through statistical exercises.