An important input when calculating sample size.

The margin of error summarizes the amount of error in the sampling process.  While the margin of error is related to the confidence interval it is NOT the same thing.  So while you might think that in our example a 95% confidence interval translates into a 5% margin of error, you would be WRONG.

Returning to the our example.  We test two executions (A vs. B) and observe a response rate of 1.20% for Execution A.  The confidence interval around A ranges from a high of 1.29% to a low of 1.11% at the 95% level of confidence. 

The margin of error here is +/- .9%.  You can see this by taking the high value (1.29%) and subtracting from it the expected value (1.20%).  Similarly, take the low value (1.11%) and subtract from it the expected value (1.20%).