How to find precision


Unfortunately, you’re unlikely to see precision reported in any of the studies in your systematic review. That’s because academics often prefer to report variance instead. Luckily, it’s easy to calculate precision using variance, because variance is just the formula for precision turned upside down.

If you remember, Precision = N/S2

And it turns out Variance=S2/N

Expressed a different way, Precision= 1/Variance.

That’s why precision is often called “inverse variance.”

You might wonder how this is possible. Earlier in the course, we said that variance = S2.
How can it also = S2/N?

The problem is, mathematicians use the word “variance” to describe two different things.

Variance can mean:

       -Variance for an individual, which is S2

       -The variance of an average of individuals, which is S2/N

Any variance that is related to the final results of a study is the second type of variance, the one describing the average of individuals, where variance = S2/N

If a study doesn’t include information about variance for its results, you can also use simple calculations to find precision using standard error or the confidence interval.

If you don’t know how to transform these numbers into precision, don’t worry! You won’t need to know these formulas if you have a statistician working with you or you have access to a computer program that can deal with different types of inputs.