Can ANYONE prove that they are using a random sample in social science/market research?
This was selected as Best Answer It is possible to do, but it’s very rare. I could, for instance, achieve a true probability sample of my social network or a small segment of people that have high affinity for a company or brand. It would be relatively easy to select a random sample, give them a fair incentive, and get 100% of the people in my sample to participate. Having said that, if one does not achieve a 100% response rate from the random sample they select, it is not really a random sample. So, for large populations, it’s often impossible to achieve a truly random sample. As you know, that’s why analysts use quota sampling, weighting and other controls.
