2

if there is no plausible sampling frame available for surveying, how will one obtain a representative sample of size n.

In the textbooks they use SRS, stratified, cluster, and systematic but these all imply a sampling frame is available like a telephone book etc.

So say for example, there is a website selling old cars, but they don't allow you to list them all, but you can search by make or model. How would you in this case get a representative sample of size n?

Or if we were trying to get a representative sample of size n to survey the number of library books that are borrowed but we don't have access to the entire catalogue?

But representative sample in the sense we can use it to make inferences about the population.

DJ_
  • 743
  • 3
  • 12
  • 23
  • 1
    I think this is a good question. However, could you please be more specific as to what you mean by representativity? For example, do you mean according to some preselected criteria, or as in a mirror image of the population, or are you mostly interested in how to draw a random sample without having a list of the whole population? – Momo Sep 25 '13 at 07:52
  • Hmm, when I first posted the question I think I was more interested in how to draw a random sample without having a list of the whole population. But representative sample in the sense we can use it to make inferences about the population. – DJ_ Sep 25 '13 at 08:00
  • Thanks. Perhaps you could edit your question to give that extra information on what you mean by "representative". You might also get some good intuition for that by browsing the sampling tag here at CV, e.g., http://stats.stackexchange.com/questions/21658/is-it-necessary-for-a-sample-to-broadly-reflect-the-target-population-distributi or http://stats.stackexchange.com/questions/61346/how-to-prove-that-selected-sample-is-representative-of-the-whole-population or http://stats.stackexchange.com/questions/18929/can-one-force-randomness-in-a-sample/ – Momo Sep 25 '13 at 08:28

0 Answers0