Random Sample - How to select?

Home > Statistics Every Writer Should Know > The Stats Board > Discusssion

Random Sample - How to select?
Message posted by C Conny on January 5, 2000 at 12:00 AM (ET)

Our company has 10,000 customers. We want to generate a random sample of 500 for a survey.

I'm not statistics savvy, so my question is, How do we go about selecting the 500 as a random sample? Do we select every 20th customer (10,000 / 500)? Do we flip a coin, starting with the first customer and then sequentially moving through the list alphabetically, and keep only those that get a "heads" on the coin toss, until we reach 500? Do we just put up a printout of all 10,000 on a big wall and throw 500 darts?

I'm not interested in theory at this point, just the way to get 500 samples and have a 95% confidence level that each was randomly selected.

Any ideas? Thanks 500 times!

C Conny

READERS RESPOND:
(In chronological order. Most recent at the bottom.)

Re: Random Sample - How to select?
Message posted by Bill on January 7, 2000 at 12:00 AM (ET)

Taking every 20th customer from a population of 10,000 (sorted alphabetically) will probably give you a representative sample. This process is called systematic sampling and works best when N is very large, as in your case.

Your $5 contribution helps cover part the $500 annual cost of keeping this site online.