91Ó°ÊÓ

A one-sided hypothesis is a type of hypothesis test that looks for an effect in one specific direction. This means that we are either only interested in detecting an increase or a decrease. For instance, if a pharmaceutical company wants to know if more patients feel headache relief with a new drug compared to a placebo, they would test to see if the relief exceeds a certain percentage, not just any change.

In scenario (c) from the exercise, the alternative hypothesis is one-sided, since the company is interested in knowing if the relief rate is more than 22%. Similarly, in scenario (d), management wants to know if the improvements have raised the acceptance rate above 60%, specifying the direction of improvement as an increase.

The one-sided hypothesis is simpler when you have a clear expectation about the direction of change. However, it limits the testing to just one direction, potentially missing changes in the opposite direction.

A two-sided hypothesis test is used when the effect or change could occur in either direction. This means the test will show significance if there's either an increase or a decrease. Two-sided tests are often more conservative because they require more evidence to reject the null hypothesis in favor of the alternative.

In the exercise from the textbook, both scenarios (a) and (b) use a two-sided hypothesis. The college dining service wants to know if there's any preference for cutlery, whether it's plastic or metal, thus not specifying a direction. Similarly, the dean's office wants to see if the study abroad application rate has changed from 10%, without indicating if they expect an increase or decrease.

When you're unsure of the direction of the effect or when both directions are important for your research, a two-sided hypothesis is preferable.

The null hypothesis, often denoted by \( H_0 \), is the statement that there is no effect or no difference. This is the hypothesis that the researcher seeks to test against. It acts as a default or starting assumption in statistical testing.

In the context of the original exercise, each scenario starts with a null hypothesis. For instance, in scenario (a), the null hypothesis is that there is no preference between plastic and metal cutlery. In scenario (b), it posits that the percentage of juniors applying for study abroad remains 10%.

The null hypothesis serves as a benchmark. A hypothesis test assesses whether the observed data provide strong enough evidence against the null hypothesis, leading to its rejection. If not, we fail to reject it, suggesting that the data do not show a statistically significant effect.

The alternative hypothesis, denoted by \( H_a \), proposes a new outcome or trend that differs from the null hypothesis. The alternative hypothesis suggests that there is an actual effect or difference that the researcher is testing for.

In the exercise examples, the alternative hypothesis gives direction to the investigation. For instance, in situation (c), the alternative hypothesis is that more than 22% of patients experience headache relief, suggesting a positive effect of the new drug. For scenario (d), it's posited that production improvements have increased the passing rate above 60%.

The alternative hypothesis is what researchers aim to support through their test results. If the data significantly deviate from what we'd expect if the null were true, the data may lend support to the alternative hypothesis being a more accurate reflection of reality.