91Ó°ÊÓ

Hypothesis testing is a statistical method used to make decisions or inferences about a population parameter based on sample data. It often involves two key hypotheses. The **null hypothesis** (denoted as \( H_0 \)) generally represents the "no effect" or "status quo" stance, suggesting that there is no significant difference or effect. In our example with the football pressures, the null hypothesis might state that the footballs used by the Patriots are not significantly underinflated from the allowable limit of 12.5 psi.

On the other hand, the **alternative hypothesis** (denoted as \( H_a \)) proposes what we seek to provide evidence for – often indicating a significant effect or difference. Here, it would suggest that the average air pressure is indeed less than 12.5 psi.

The outcome of a hypothesis test is determined by the test statistic, which is calculated from the sample data. This value helps in deciding whether to reject or not reject the null hypothesis. The probability of observing a test statistic as extreme as, or more extreme than, the observed value is captured by the **p-value**. If this p-value is less than a pre-determined significance level (commonly 0.05), we reject the null hypothesis, indicating statistical significance.

The t-distribution is a crucial concept encountered when performing hypothesis testing, especially when the sample size is small or the population standard deviation is unknown. It resembles the normal distribution but has heavier tails, which provides for the additional variability inherent with smaller samples. This makes it particularly useful in contexts where sample sizes are under 30, as is the case in our example with 11 footballs.

The shape of the t-distribution is determined by the **degrees of freedom**, typically calculated as the sample size minus one (\( n - 1 \)). For our data, this results in 10 degrees of freedom. With more data points, the t-distribution approaches a normal distribution. That's why for larger sample sizes, we often use the normal distribution (z-distribution).

Understanding the t-distribution assists in calculating the test statistic and p-value, both essential in comparing against the null hypothesis. With heavy tails, the t-distribution accommodates greater variability, potentially affecting decisions about the null hypothesis based on sample data.

Statistical significance is a determination that results from a hypothesis test, indicating whether an effect is likely not due to random chance. It allows researchers to decide whether to reject the null hypothesis based on sample data.

In our football pressure example, we derived a p-value less than the commonly accepted significance level of 0.05, suggesting the results were statistically significant. This means the observed sample mean of 11.1 psi is unlikely to have happened by random chance, providing evidence against the null hypothesis that the true mean matches the allowable limit.

However, statistical significance isn't about the magnitude or practical importance of a difference. A result can be statistically significant but may not be practically important. This highlights the benefit of pairing statistical significance with effect size and other practical considerations for a complete perspective on data findings. Understanding this helps in weighing the implications of results beyond mere statistical computations.