Channelling Nero Wolfe?

[bcholmes]

"Contact is not a verb."
"Yes sir, it is. Last time you corrected me I went home and looked it up. Transitive and intransitive."
"Not in this house, it isn't!"

A coupl'a times I've watched people quibble about the difference between the words "precision" and "accuracy". I've never really felt that their definitions were different.

I popped over to dictionary.com today and found that the second definition of accuracy is "Precision; exactness". Hurm.

Comments

Precision and accuracy.

[baratron.livejournal.com]

There's a difference in maths/science. "Precision" refers to the number of significant figures given. Stating that something measures 0.500000mm implies that it is as close to exactly 0.5mm as you can measure. As scientists, we are taught not to quote calculations to more significant figures than can actually be measured. If the smallest graduation on your burette is 0.2ml, you should bear that in mind when calculating the volume of liquid used - you can't say you used 430.400ml because you can't measure that precisely.

"Accuracy" refers to how certain you are that a figure is correct. So you may be able to quote your result with 3 degrees of precision (say it's 0.487 in some unit), but is that accurate? If the bottle is supposed to hold 0.5 of that unit, then 0.487, even though you've measured it precisely, is hardly accurate. I hope this explanation makes sense!

I agree with you when it comes to ordinary English, though. I have the same problem with people who insist that jealousy and envy are different. Maybe for them, the two words describe different emotions, but in my lexicon, they're different words for the same thing. Simply changing the name you put on something doesn't change how it feels in my heart. But I have digressed wildly...

Re: Precision and accuracy.

[purplesofa.livejournal.com]

I concur with baratron's explanation.

To the layperson, "precise" and "accurate" are prety much interchangeable. But scientists and statisticians have applied more (ahem) precise meanings to the terms.

In statistics, we would say that precision refers to the size of the "confidence interval". E.g. if you do some sampling to estimate that the average Toronto subway car contains 150 people, you might be off by plus or minus 20 people. That's the level of precision, and it can be computed by statistical methods.

In statistics, the opposite of accuracy is bias. If you had sampled the subway car populations only at rush-hour, you would get an upward bias, and thus an inaccurate estimate. Bias can't usually be computed - or else it would have been prevented in the first place, by more accurate measurement methods.