Researchers are investigating whether people who exercise with a training partner have a greater increase, on average, in targeted exercise intensity compared with people who exercise alone. Two methods of collecting data have been proposed.

Method I: Recruit volunteers who are willing to participate. Randomly assign each participant to exercise with a training partner or to exercise alone.
Method II: Select a random sample of people from all the people who exercise at a community fitness center. Ask each person in the sample whether they use a training partner, and use the response to create the two groups.

Suppose the investigation produces a result that is statistically significant using both methods. What can be concluded if method I is used that cannot be concluded if method II is used? Explain your answer.

The difference between the two experiments described is in the mode of assignment of the participants to either of the two groups. Hence, the use of random assignment in method 1 will allow for a more generalized conclusion to be inferred as it would perform better on highly variable samples compared to method 2.

The result of method 2 would work better at making inferences when the type of training regime which the person undertakes is known while method 1 will be more preferred for making inferences about samples without a known or established training regime.

Therefore, it can be concluded that method 1 would perform better on diverse or highly variable samples than method 1 would.

Learn more :link

The conclusion that you will be able to reach if method I is used that cannot be concluded if method II is used is that the results can be generalized to the population as a whole.

In method I, the researchers decided to recruit volunteers. This sample includes people who do not exercise, people who exercise alone and people who exercise with a partner. On the other hand, method II will recruit participants at a community fitness center. Therefore, this sample will not include people who do not exercise. This means that method I will have a sample that is more representative of the general population.