A percentage estimate (i.e. 95%) of how confident we are that the actual value lies within the range (interval) of our estimated value.
Example: Imagine a hypothetical company that provides a super low-cost internet search tool. Let’s call them Froogle. Let’s estimate the average salary of the engineers working at Froogle.
We ask 1o0 Froogle engineers for their salary, and find that the average is $50k. However, those are only 1o0 engineers in a company of 1,000 engineers. And remember, we’re trying to find the average salary of all Froogle engineers.
Since we only asked 10 engineers, we can’t be too sure the average salary actually is $50k. Using statistical analysis we come up with ranges and confidence intervals. For example, we may estimate that the actual salary is between $35k-$75k with a confidence interval of 95%. I.e. we are 95% sure that the actual average salary falls somewhere between $35k and $75k.
Wisdom: “There are three types of lies—lies, damn lies, and statistics.” ― Benjamin Disraeli