91Ó°ÊÓ

Hypothesis testing in statistics begins with two contrasting statements - the null hypothesis (\(H_0\)) and the alternative hypothesis (\(H_1\text{ or }H_a\)). The null hypothesis represents a default position that there is no effect or no difference and is typically the statement being tested. For example, in the context of The Flipper machine, the null hypothesis is that it flips heads at least 88% of the time, or \(p_0 = 0.88\).

The alternative hypothesis, on the other hand, challenges the null hypothesis by proposing what we suspect might be true instead. It is a claim that must be tested. In our coin-flipping scenario, the alternative hypothesis suggests that The Flipper lands heads less than 88% of the time, expressed as \(p < 0.88\).

It's vital to establish these hypotheses correctly as they formulate the basis on which the validity of the claim will be tested, guiding the direction of statistical testing. In hypothesis testing, we aim to collect evidence to either reject the null hypothesis in favor of the alternative or to fail to reject the null hypothesis, which would imply there is not enough evidence against it.

The sample proportion is a statistic that estimates the proportion of successes in a population based on a sample. It is denoted as \(\hat{p}\) and is calculated by dividing the number of successes by the total number of observations in the sample. In the context of our example, flipping a head is considered success. If The Flipper lands 172 heads out of 200 flips, the sample proportion would be \(\hat{p} = \frac{172}{200} = 0.86\).

This figure provides a snapshot estimate of the true proportion that would be obtained if the entire population could be studied. When we have a large enough sample size, the sample proportion can be assumed to be approximately normally distributed around the true population proportion—this assumption underpins our ability to perform hypothesis testing using the sample proportion.

The z-value, or z-score, plays a critical role in hypothesis testing. It is a measure of how many standard deviations away a data point is from the mean. To calculate the z-score for sample proportions, we use the formula \(z = (\hat{p} - p_0) / \sqrt{(p_0(1 - p_0)/n)}\), where \(\hat{p}\) is the sample proportion, \(p_0\) is the hypothesized population proportion under the null hypothesis, and \(n\) is the sample size. The resulting z-value tells us whether the observed sample proportion is significantly different from the null hypothesis proportion.

For instance, if scenario (b) resulted in a z-value higher in absolute value than the critical z-value, this would mean the sample proportion is significantly lower than the claimed 88%, leading us to consider rejecting the null hypothesis.

The significance level, denoted as alpha (\(\alpha\)), is the threshold that determines whether a result is statistically significant. It represents the probability of rejecting the null hypothesis when it is actually true—the risk we are willing to take for a false positive. A common choice for \(\alpha\) is 0.05, meaning there is a 5% chance of concluding that there is an effect when there is none.

The significance level aids in determining the critical value for a hypothesis test, which then guides us in deciding whether to reject the null hypothesis. When comparing our calculated z-value to the critical z-value informed by our \(\alpha\), we are looking to see if the evidence is strong enough to conclude that the sample provides sufficient evidence against the null hypothesis, within the acceptable risk of error defined by the significance level.