rkillings May 16, 2011:

dispersion If what you are grasping for is the name of the phenomenon that is *measured* by standard deviation, it's this. Specifically, dispersion about a measure of central tendency.

B D Finch May 16, 2011:

"Type"! "Discrepancy" would be too vague and not appropriate as it is not a statistical term as far as I am aware. "Variation" in statistics is used for the sum of the squared deviations from the mean. The term used in the source text is not "écart", but "écart-type", so I cannot see any reason for not translating it as "standard deviation", which also does make sense in the context.

kashew (asker) May 16, 2011:

Variation and discrepancy are running around in my brain. What do you think?

B D Finch May 15, 2011:

@rkillings Thanks for the reminder that the earth is round and paper and computer screens are flat!

rkillings May 15, 2011:

planimetry Any mapping of the earth's surface to a plane is an *inherently* statistical exercise. At best, the standard deviation of any point estimate is a few centimetres.

B D Finch May 15, 2011:

Statistical? I think that it is not so obvious how SD is relevant here. It cannot be about the difference between two points. "... d’un point quelconque du canevas primaire par rapport au canevas national There can be differences in the projection of the two "canevas", as well as inaccuracies or even changes that may have occurred due to geological movement of land or human activity. In establishing the position of the "point quelconque", at least three reference points on the "canevas national" will be used. When e.g. staking out a footprint on the basis of the first point one then checks a different point on the "canevas primaire" it may be found that the geometry is inconsistent. I think this is where SDs come in. I have, incidentally, come across a very similar problem with setting out a new development.

Catharine Cellier-Smart May 15, 2011:

I thought you'd know the normal translation which is why I was surprised by your question. But FWIW "standard deviation" gets many hits in the context of "land survey"; see for example http://www.surveyors.com/land-surveyors/land-survey-formulas...

kashew (asker) May 15, 2011:

@ C C-M That's the raison d'être of my question! Isn't it non-statistical i.e. simpler, for a variation or discrepancy between a couple of measurements?
I do know the normal translation.

Bourth (X) May 15, 2011:

Kashew obviously has a sexual aversion to deviation in any shape or form.

Evans (X) May 15, 2011:

absolutely agree!

polyglot45 May 15, 2011:

standard deviation THE term

Hilary Wilson May 15, 2011:

Indeed!

Catharine Cellier-Smart May 15, 2011:

Any particular reason you don't want to use the term "standard deviation" ?