So how can we collect sample data of average valuation of US start up company that is random and representative? Could collecting data from any start up company from each states be considered random and representative?

Hi Nicholas, Technically convenience sampling is a non-probability technique. It is not highly recommended due to possible sampling errors due to possible misrepresentation of some parts of the true population. But I'm not sure if I can recommend a better way to collect this sample if you don't work with highly expensive databases that give you access to more representative information. Best, Ned

Hi Nicholas, You can use the so-called 'convenience sampling', which allows you to draw observations from that part of the population, which is close at hand. However, for your results not to be biased you need to make sure there isn't a pattern whatsoever in which you select which parts of the population will be included in your analysis (or if you do, you need to be aware of that). Best, Ned

Thank you, Ned! Does that mean convinience sampling is a type of non-random sampling since we're only collecting data close at hand? If so, wouldn't it be impossible for that data to be fully representative to the population being studied?