91Ó°ÊÓ

The P-value is an essential tool in hypothesis testing as it helps us decide whether to reject the null hypothesis. It measures the probability of obtaining test results at least as extreme as the ones observed, assuming the null hypothesis is true. In our example, the P-value is a probability related to the t-distribution. When we calculated the t-statistic to be approximately -3.71, we wanted to determine how extreme this result was under the assumption that the null hypothesis (\(\mu = 5\)) is correct. We did this by looking it up in a t-distribution table or using a calculator.Since the test is two-tailed, we must consider both tails of the distribution. By doing this, we found a P-value of approximately 0.0003. This small P-value indicates that observing such extreme values would be very improbable if the null hypothesis were true. That's why we compared the P-value to our significance level, \(\alpha = 0.01\), and since 0.0003 < 0.01, we rejected the null hypothesis.

The t-distribution is a key component of hypothesis testing, especially when dealing with small sample sizes. It's similar to a normal distribution but has thicker tails, making it more prone to yield values far from its mean. This characteristic is helpful when you do not know the population standard deviation and need to work with the sample standard deviation instead. For this exercise, we used the t-distribution to calculate the test statistic because we aimed to determine how far our sample mean (\(\bar{x}=4.87\)) was from the claimed population mean (\(\mu=5\)), given the sample standard deviation and sample size. By substituting the given values into the t-test statistic formula, we found a value of -3.71. This indicates the number of standard errors that the sample mean is away from the hypothesized population mean. When using 99 degrees of freedom (\(n-1=100-1\)), the t-distribution allows us to find the P-value associated with our test statistic.

The null hypothesis is a fundamental concept in hypothesis testing. It represents the default or "no effect" assumption, stating that there is no difference or change. In this case, the null hypothesis (\(H_0: \mu = 5\)) suggests the mean weight per tablet is exactly 5 grains, as advertised.The purpose of setting up a null hypothesis is to provide a baseline against which we can compare our sample data. Under the null hypothesis assumption, we calculate probabilities to decide whether the data we have is consistent with this hypothesis. If our test results show that such data is highly unlikely, given the null hypothesis is true, we reject it. Otherwise, we fail to reject it and generally accept the null hypothesis. Rejecting the null hypothesis in our aspirin weight example suggests that the actual mean weight differs from the claimed 5 grains.

The alternative hypothesis acts as a counterpoint to the null hypothesis in hypothesis testing. It proposes what we suspect might be true instead of the null hypothesis. In the aspirin exercise, the alternative hypothesis (\(H_a: \mu eq 5\)) suggests that the mean weight per tablet is not equal to 5 grains.Selecting the correct form of the alternative hypothesis is dependent on the research question. Here, we opt for a two-tailed test because we want to know if the weight is either more or less than the advertised 5 grains, not just one over the other. By rejecting the null hypothesis in favor of the alternative hypothesis, our analysis provides strong evidence suggesting that the tablets' mean weight does indeed differ from the claimed 5 grains, hinting that the manufacturer might need to revisit their filling process.