Tuesday, December 8, 2009

Wondering how big of a sample size you need?

So, you have decided to do an evaluation - or are doing some preliminary research for a proposed program. You sit down and try to figure out the details, which includes how many people should you send your survey to? If you are just doing asking a bunch of people to participate in a survey asking about why they donate, or whether they enjoyed your program, then the sample size probably isn't as important. It is important is you want to be able to generalize your findings to the general population - or to the targeted population. So, how do you determine it? Well I could tell you the complex formula and math behind determining a sample size, but it is easier just to point you to a simple sample size calculator you can download. This link will take you to a survey course website, on the bottom left of the page you will see "Sample Size Calculator" click it and download.

Once it pops up, it might be a little confusing so here are a few tips to make it easier:

  • The first tab "Type of Analysis" you can usually leave the defaults - unless you are doing complex sampling - which you probably aren't.
  • The second tab "Values and Settings" is most important. Make sure to enter your population size, etc.
  • The third tab "Corrections" is pretty much self-explanatory and you will probably not use it - but if you do it explains what each option means by the selection box.
  • Once you have entered everything in, then the box to the right should say a number - that is the number of people your sample should include.
*When you download it, there is a "quickhelp" folder that explains what each box means if you are confused about what to put there. Good luck!


Alexander said...

Good post very interesting