why use random sample?

Answer 1

Well, you don't say as opposed to what, but I'll give it a try.

Selection of a random sample allows for inferences to be made back to the larger population from which it came. Using other nonprobability methods do not allow for this. There ARE other sampling methods besides simple random that would also allow for inference back to the population. These include: systematic, stratified, cluster, etc.

Answer 2

It's important to use random sampling because you want the closest representation possible to the general population. For example, if you are doing a study on the effects of caffeine on grades and you decided to use 2 psychology classes for your subject pool. You decide that the 8:00 a.m. class will be your control group (receive no caffeine before the test) and the 11:00 a.m. class will be the experimental group (they each drink a 16 oz. coffee before taking their test). Your results may not be valid. The 8 a.m. class may possess characteristics that the 11 a.m. class doesn't. They may drink more caffeine before leaving the house early, they be more motivated students, they may have a different teacher. There are many factors that could alter the results. But if you use random sampling in both classes to get your participants you are more likely to get a mix of people in each group. You could hand out numbers and all even numbers from both classes are the control group while the odd numbers are the experimental group. That way every participant has an equal chance of being in either group. There's more to it but that would take much more time to explain.