Tuesday, April 30, 2013

Margin of Error - Here's a quick rule of thumb to the question: "What's the best sample size?"

An often asked question is "How many responses do I need to get on my survey to make sure that it is valid?"

Jeff Tranguch, Research Analyst with MGI, prepared a very quick email to one of my clients on precisely this today. A really bright guy, I wanted to share it with you because of its simplicity and brevity.

Here it is in his words...

"As you can see, the number drops as you approach 1000 responses. They will have some direction with about 500 or so responses (obviously they would have better direction with 1,000 or so responses).  

All of the numbers below are calculated at a 95% confidence interval which basically means that if we were to conduct this study 100 times, the same results, plus or minus the margin of error would occur 95 out of 100 times."

Number of Responses
Margin of Error
100
+/-9.8%
250
+/-6.2%
500
+/-4.4%
750
+/-3.6%
1000
+/-3.1%
1500
+/-2.5%



If you have any questions, please feel free to call or email me.

1 comment: