Take the 2-minute tour ×
Physics Stack Exchange is a question and answer site for active researchers, academics and students of physics. It's 100% free, no registration required.

Accuracy and precision are among the most fundamental concepts in experimental physics, and, I always believed, completely unambiguous.

Recently I found that the Wikipedia article on Accuracy and Precision claims that a "shift in the meaning of these terms" is occurring. My first thought was that this must be a joke, a mistake or Wiki vandalism.

However, the ISO 5725-1 (Accuracy (trueness and precision) of measurement methods and results) standard referenced in the article indeed reads

3.7
accuracy

[...]

NOTE 3 Accuracy refers to a combination of trueness and precision.

Notably, this seems to imply that the concept of precision is a subset of the concept of accuracy, which is plainly incompatible with the definitions found almost everywhere else that sharply contrast precision and accuracy.


My questions are:

  1. Is such a "shift in meaning" really occurring in the physics/measurement community at large?
  2. If yes, what is the cause of this change of terminology?
share|improve this question
2  
This question appears to be off-topic because it is about language usage and not physics or physics concepts. –  Brandon Enright Jul 17 at 6:46
7  
It's not about language but about terminology and definitions, both of which have tags on this site with hundreds of questions each. –  pew Jul 17 at 6:59
1  
Your question "is the change really occurring" is a question that requires linguistics data, not physics data. Your question "what is the cause" requires linguistic analysis and sociology. –  Brandon Enright Jul 17 at 7:01
2  
No, Brandon, you're wrong. The "shift in the meaning" isn't a claim about linguistics data - usage of language by the physicists or other parts of the public - but a statement about a particular document, ISO 5725, which decided to interpret the terms a bit differently. –  Luboš Motl Jul 17 at 9:23
2  
I agree that this is an important topic -- just look at the hairball of politics caused by the public's misunderstanding of the scientific term "theory of X." IMHO the last thing we need is some ISO doc fouling up the meaning of other terminologies. –  Carl Witthoft Jul 17 at 12:01

3 Answers 3

up vote 9 down vote accepted

The "shift in the meaning" refers to some attempts to reinterpret the terminology that were made by a metrological document, ISO 5725, in 2008. That may be described as a bureaucratic effort by a few officials – really bureaucrats of a sort – and as far as I know, the "shift in the meaning" hasn't penetrated to the community of professionals. The people behind the standards have certain limited tools to assure that their new interpretation of the words would be taught to students of physics and engineering but I don't think that they are succeeding so far.

Before 2008, people would agree that "precision" refers to the typical difference between individual measurements, and the precision is good if a "statistical error" is low or if the measurements are producing "many significant figures" for the result.

"Precision" didn't – and still doesn't – discuss whether the results of the measurements are actually clustered around the true value. They may be separated by a "systematic error" – which goes in the same direction in each repetition of the measurement and can't be removed by averaging many measurements. The absence (or low magnitude) of this "true" systematic error – the difference between the true value and the average of many measurements – would be summarized by the word "accuracy".

One wants results that are both "precise" and "accurate" in the sense above, and a word like "valid" or "satisfactory" or another similarly neutral non-technical word would be reserved for results that are both "precise" and "accurate".

The "shifted" 2008 proposal is to use the word "accurate" for what would previously be called "valid" etc. – i.e. for measurements that suppress both kinds of errors, systematic and statistical, i.e. for measurements that are both accurate (in the pre-2008 sense) and precise (in the pre-2008 sense which is the same as post-2008 sense).

Even if this "shift" succeeded in the language of professionals, it won't make much difference. The reason is that the word "accurate" has pretty much implicitly included "precise" even before 2008. If you make a small number of measurements (repetitions of a measurement) and you want to determine whether the measurements are "accurate", you have to calculate the difference between the true value and the measured values. But if you have just one or several measured values, the difference is affected by the "statistical error", anyway, so you can't quantify the "systematic error" well, anyway. To be sure that the systematic error is low, after just a few measurements, the statistical error has to be low, too (the precision has to be good).

So whether the word "accuracy" included "precision" before 2008 is debatable. All these things are just changes at the level of the language. When one is quantitative, things have to be described by actual quantities – systematic errors and statistical errors and nothing was changed about the meaning of these quantities in the 2008 document.

share|improve this answer

1. No, there is no "shift in meaning".

"Accuracy", "precision", and "trueness" is a technical term for measurement not physics. And there is no such thing as a "measurement community" because measurement occurs everywhere. As such, "accuracy", "precision", and "trueness" are heavily overloaded technical terms used in varying fields like maths, computer science, engineering, statistics, etc.

It goes without saying that the white-collars of different fields thus have different ideas of what "accuracy", "precision", and "trueness" mean / should mean. In practice, it's best to first define the terms upfront before using them in writing or speech (cf. NDT's example) because unlike certain privileged technical words that may have that one-true-meaning, these three words are notorious for their lack of consensus.

2. No, it's unlikely that the meaning of "accuracy", "precision", and "trueness" will ever change in the near future.

Trying to change the meaning of a terminology of a particular field or subject may be a manageable task, However, as much effort as ISO 5725-1 has put in, it is still not conceivable that the meaning of a terminology overloaded by so many different fields can be changed at large.

3. Regarding your quote "concept of precision is a subset of the concept of accuracy"...

Since definitions differ, depending on how you define "accuracy" and "precision", the concept of precision may be included in the concept of accuracy. If "accuracy" means that the "average measured value" is near the "true value", and "precision" means that the "measured values" are near one another, then accuracy requires that the "average measured value" take only precise values as input.

In other words, without first having precision, we wouldn't even have an "average measured value" to speak of:

enter image description here enter image description here enter image description here

share|improve this answer
    
-1. Your use of accuracy is precisely the use that the BIPM (and ISO, following BIPM) want to eliminate. You have ignored the gist of the question. There most certainly is a shift in meaning in metrology. The BIPM wanted a single word that (a) incorporates the concepts of systematic and random error, and (b) has a clear and simple translation between English and French. At least within the metrology community, there most certainly is a shift in meaning. –  David Hammen Sep 11 at 11:51
    
@DavidHammen, That's ok, but your comment proves that my answer is correct. As I've mentioned, metrology (science of measurement) is not the only field doing measuring. Everyone is doing measurements and the concepts of systematic and random error do not only exist within metrology; they have been addressed in other fields as well..................................... –  Pacerier Sep 11 at 12:06
    
...............It's true that the BIPM wanted "(a)" and "(b)"—in fact, no one said that that's untrue—, but look at the question again: quote "Is shift in meaning really occurring in the physics/measurement community at large?" unquote. Answer is No, shift in meaning is not really occuring in the "measurement community" at large. ("measurement community" means the whole world, "measurement community at large" also means the whole world) –  Pacerier Sep 11 at 12:06

I'm not familiar with ISO 5725 (a 1994 revision of a 1986 document, apparently "reviewed and confirmed" in 2012), and it seems that I have to buy it to read it. A 2008 vocabulary of metrology put out by the BIPM and also cited by Wikipedia has definitions much closer to my intuition, and to common usage among folks I know who specialize in precision measurements:

2.13(3.5)
measurement accuracy
accuracy of measurement
accuracy

closeness of agreement between a measured quantity value and a true quantity value of a measurand

NOTE 2 The term “measurement accuracy” should not be used for measurement trueness and the term measurement precision should not be used for ‘measurement accuracy’, which, however, is related to both these concepts.

2.14
measurement trueness
trueness of measurement
trueness

closeness of agreement between the average of an infinite number of replicate measured quantity values and a reference quantity value

NOTE 3 Measurement accuracy should not be used for ‘measurement trueness’ and vice versa.

2.15
measurement precision
precision
closeness of agreement between indications or measured quantity values obtained by replicate measurements on the same or similar objects under specified conditions

NOTE 4 Sometimes “measurement precision” is erroneously used to mean measurement accuracy.

However, it's worth remembering that in the extreme of terrible precision, so many measurements may be needed to determine whether the average is accurate or not that the measurements may be effectively useless. Remember the old joke about the three economists playing darts at the pub: the first puts his darts in the wall above the dartboard, the second puts his darts in the wall below the dartboard, and third gets excited and yells "Bullseye!"

share|improve this answer
    
-1. The BIPM's wording is exactly what motivated the change to the ISO 5725. The BIPM specifically says that accuracy encompasses both systematic error (trueness) and random error (precision). –  David Hammen Sep 11 at 11:53
    
Shifts in meaning don't occur by the decree of a standards organization. Next time I'm at a conference with metrology folks or fundamental constants folks I'll bring up the subject. –  rob Sep 12 at 4:35
    
Sure it does. We physicists have tried to do exactly the same thing with "weigh" and "weight". Do you criticize your friends who say they weigh a certain number of kilograms? Your answer is rather hypocritical if you have. –  David Hammen Sep 12 at 4:38

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.