Question

Lucy recently asked the servers at her restaurant to only give straws to customers who request them. She thinks that about half of the customers will ask for straws but hopes that the rate will be less than half. She randomly selects 100 customers and finds that 43 of them ask for a straw. To determine if these data provide convincing evidence that the proportion of customers who will ask for a straw is less than 50%, 150 trials of a simulation are conducted. Lucy is testing the hypotheses: H0: p = 50% and Ha: p < 50%, where p = the true proportion of customers who will ask for a straw. Based on the results of the simulation, the estimated P-value is 0.0733. Using ∝ = 0.05, what conclusion should Lucy reach?

Because the P-value of 0.0733 > Alpha, Lucy should reject Ha. There is convincing evidence that the proportion of customers who will ask for a straw is less than 50%.
Because the P-value of 0.0733 > Alpha, Lucy should reject Ha. There is not convincing evidence that the proportion of customers who will ask for a straw is less than 50%.
Because the P-value of 0.0733 > Alpha, Lucy should fail to reject H0. There is convincing evidence that the proportion of customers who will ask for a straw is less than 50%.
Because the P-value of 0.0733 > Alpha, Lucy should fail to reject H0. There is not convincing evidence that the proportion of customers who will ask for a straw is less than 50%.

Answer 1

Because the P-value of 0.0733 > Alpha, Lucy should fail to reject H0. There is not convincing evidence that the proportion of customers who will ask for a straw is less than 50%.

Explanation:

To understand this question, you'll have to understand a few terms: the null hypothesis, the alternative hypothesis, the p-value and the signifiance level.

The null hypothesis (H₀) is the assumption that the phenomenon doesn't exist - in this case, it is that the proportion of customers who will ask for a straw is 50%, or due to chance (you can either ask for a straw or not). When doing a statistical test for the existance of a phenomenon, we usually assume the null hypothesis to be true first.

The alternative hypothesis (Hₐ) is the assumption that a phenomenon is true. In this case, it is that that the proportion of customers who will ask for a straw is less than 50%.

The p-value is usually a result you get after performing the statistical test - it's the probability that the observed results are at least as extreme as the results actually observed, assuming that the null hypothesis is true. In this case, assuming that the proportion of customers who will ask for a straw is 50% (the null hypothesis), a p-value of 0.0733 for the observed result of 43 customers means that if the experiment were to be repeated again, then there is a 7.33% chance that 43 customers or less will ask for a straw, assuming the null hypothesis is true. This seems pretty good - assuming the null hypothesis is true, there's only a 7.33% chance that we could have gotten 43 customers or less, which means it may be reasonable to "reject the null hypothesis" (i.e. the alternative hypothesis may be true).

However, in many cases a 7.33% chance may not be low enough, e.g. in this problem, a less than 5% chance (i.e. p < 0.05) is necessary. This required chance for results to be statistically significantly is called the signifiance level, or alpha value, the "the probability of rejecting the null hypothesis when the null hypothesis is in fact true." In scientific journals or depending on the field of study, there may be a generally accepted signifiance level.