What’s the difference and why – and when – do I care? Let’s start by defining and
differentiating the two terms. If you’ve ever picked up a bow, then this graphic will make
sense:
Precision relates to tightness of the grouping and speaks to how repeatable the measurement is.
Accuracy is how close we are to center. In this case, it’s how close are we to the actual value.
And the best measurements are in the lower right hand target.
For the mathematically (and Lean) inclined among us, you can also visualize these two in statistical
terms:
Here, Precision correlates to the magnitude of Standard Deviation, while Accuracy is
determined by the distance between the mean and reference values.
So when do I care about either? Well, as in all engineering decisions, it depends.
When developing a closed-loop process, the goal is normally to maintain a constant output so the loop
can adjust for process variation. In this case, the actual measurement value (Accuracy) isn’t as critical as
the stability and repeatability of the measurement (Precision).
On the other hand, if you are certifying performance for Energy Star compliance, you must have an NIST-
traceable measurement within a certain range. Here, Accuracy dominates – but not to the exclusion of
repeatability.
The toughest cases are billing measurements. People get a little tense when money is involved and
demand to understand a) how the measurement is made and b) why it supports the bill presented.
Think Renewable Energy sites with Battery Energy Storage Systems (BESS) owned by different parties.
Both parties will insist system accuracy be traceable to a NIST or IEC standard – AND be repeatable.
We can solve any of these by understanding Precision and Accuracy – and their Nemesis, Uncertainty
(also known as “Errors”).
Join us next time for “Uncertainty Explained”.
