Title
In: 2
Margin of error has a technical meaning when you’re talking about sampling.
Imagine you want to know how the proportion of red and black cards there are in a deck of playing cards, but you don’t want to look through every card. You shuffle the deck and draw out two cards. You get two red cards. So you estimate the deck is all red cards.
Now I think it’s obvious you haven’t drawn enough cards there to get a reliable estimate. If that were a normal deck and you pulled two random cards again (from a full deck) you’d have a 1 in 3 chance of getting the complete opposite result – two black cards!
You’d get a more reliable estimate by drawing 10 cards, and even more reliable estimate by drawing 20. If you drew all 52 cards you’d be 100% accurate.
But just *how* reliable are these estimates?
This is what the margin of error tells you. For a given level of “confidence” – most commonly 95% – it gives you a range and tells you a range – say, +/-10%. What this means is that if you repeated that sampling process multiple times, the true value would be within that range 95% of the time.
Latest Answers