Random sampling is a way of sampling data. In a random sample every element in the population has an equal chance of being selected. For instance, to do an election poll, calling people randomly out of the phone book would be better than asking people only at a labor union meeting or at a small business convention. A page of the phone book is a more random sample of voters than a union meeting, so the result is more likely to tell who will really be elected. But neither sample is really random because for one thing, some people do not have phones.