Online Surveys = Ongoing Insights

Tasting the Fruit Salad: How to Get Your Sample the Right Way

 → 
 → 
Tasting the Fruit Salad: How to Get Your Sample the Right Way
Fruit Salad

Picture the following problem…the sun is shining, the birds are tweeting, and you are at a company picnic on a Saturday. But that’s not the problem. Here’s the problem: you are standing face to face with a giant delicious looking bowl of fruit salad. Unless you want to get fired, you can’t just dump the whole bowl on your plate, so you’re going to have to spoon just a little bit of it on your plate instead. The question racing through your mind as your spoon hovers over the bowl is: “How do I choose what goes on my plate?”

Well, the world of sampling fruit salad is actually quite similar to the world of sampling survey respondents. And, as is true for so many of these blog posts, the answer is—it depends! There are two main types of sampling strategies. To know which strategy to use, first you have to know what your research question is. (Need help with that part? Check out our blog post on that here.) If you’re all set with your research question then read on…

Strategy #1: Proportional Sampling

What is it? In proportional sampling, your sample will look similar to your population in that it maintains the same proportions. So if the fruit salad is 75% apples and 25% grapes you would want your sample to be 75% apples and 25% grapes as well. For example, you might sample 3 apples and 1 grape from the giant bowl of 750 apples and 250 grapes.

When do you use it? You use proportional sampling when you want to know what the big bowl tastes like overall. To do that you have to make sure you honor the chef’s recipe and keep the same proportion of apples and grapes in the spoonful you take.

Okay, I get the fruit example, but what kind of survey would I use it for?

  • A research question that tests: On average, how much do American voters support same-sex marriage?
  • The sample you’d need: A representative sample of Democrats, Republicans, and Independents.

The bottom line? Use proportional sampling to get an overall picture. 

Strategy #2: Oversampling

What is it? In oversampling, your sample will include an equal balance of every group of interest in your population—regardless of the naturally occurring proportion in the population. So even though the salad is 75% apples and 25% grapes, you would want your sample to include 50% apples and 50% grapes—so you’d have to oversample the grapes. For example, you might sample 2 apples and 2 grapes from the giant bowl of 750 apples and 250 grapes.

When do you use it? You use oversampling when you want to compare how the apples taste to how the grapes taste. In order to do that you need to make sure you pick enough grapes out of the bowl (even though there are fewer of them there) to give each fruit an equal shot of being tasty.

Okay, I get the fruit example, but what kind of survey would I use it for?

  • A research question that tests: Do different types of American voters differ in their support for same-sex marriage?
  • The sample you’d need: A sample with equal amounts of Democrats, Republicans, and Independents.

The bottom line? Use oversampling to compare groups.

Unsure which sampling strategy to use for your research question? Post it in the comments below…

Tags: , , ,

Inspired? Create your own survey.

Inspired? Create your own survey.

PRO Sign Up Sign Up FREE or Sign in
Write Surveys Like a Pro

Ever wonder what SurveyMonkey’s really made of?

Ever wonder what SurveyMonkey's really made of?

Read our engineering blog »