Stats question

[Yume]

Question:

Do college students learn basic accounting better in a classroom or using an online course? Is this question most reasonably answered with an experiment, a sample survey, or an observational study that is not a sample survey? Briefly explain your choice. If your answer is a sample survey, determine the population and the sample.

The answer is that it's an experiment and you want to see if there's a cause-and-effect between teaching method and amount learned.

My question is, why isn't it an observational study? Wouldn't it be better to observe people who use either methods and then see which methods produce better results?

What treatments are you imposing on the students? I know the factor is Teaching method.

[Harlequin]

Now I've never taken stats so don't take this as an end-all answer, but from what I've found...

It's not observational study because the assignment of subjects to a control group versus a treated group is outside of your control. This would mean that every one you're looking at is (most likely) within their comfort zone. It's better to get an average from groups that are more randomized, correct?

[Yume]

with this explanation+other explanation I got from some friends, I understand now.

Thank you!