Safe Speed Forums

This is a split off another thread that went off topic...



Precision has nothing to do with the actual VALUE of the reading, only with the instrument's ability to make the reading to the required resolution. An instrument with a high degree of precision or resolution (0.01 +/-0.005mm in the case of vernier callipers for example) is absolutely useless if it is not accurate because it is used at the wrong temperature for examle. In this case, the instrument may READ 1.25mm when the actual VALUE is 1.55mm. In this case, you have a fairly high degree of precision but low accurecy. However, an instrument with a low degree of precision or resolution (ie 1 +/-0.5mm for a steel rule) used at 22 deg. C may read 1mm when the true value is 1mm. In this case you have a low precision with high accuracy. It is more difficult to obtain high accuracy in a high precision instrument, as the higher the precision the tighter the tollerance (or allowable discrepency). Therefore in order to be accurate the instrument must be callibrated more carefully. It also becomes increasingly difficult to maintain good accuracy the larger the measured value becomes, as small discrepencies in the reading are magnified. Precision is constant regardless of the value of the measurement taken.

I think you're getting bogged down with tollerance and callibration, and other system factors smeggy. Accuracy and precision are, in a simple sence, very basic concepts.

_________________
Science won over religion when they started installing lightning rods on churches.

and whats it about then ... ?

Wait for me to type it then!

_________________
Science won over religion when they started installing lightning rods on churches.

Thanks, but...

That wasn't what I said



That's what I meant!

_________________
Keep right on to the end of the road ...

Thanks for splitting the thread. I’m sure that almost everyone else was growing tired of us. However, I would prefer that a moderator moved everything from my pedantic post onwards.



I sort of agree but not entirely, I hope this is the source of our disagreement (we’ll be going in circles if not). Resolution is not a measure of precision (although a level of resolution is required to demonstrate a level of precision). You can have an instrument with a high resolution which can still be imprecise and inaccurate (due to e.g. stability).

I just had a surf and came across this (IEEE):

http://www.ieee-uffc.org/freqcontrol/qu ... gaccur.htm

In our case, the speedo described will fit into the first column: precise (repeatable) but inaccurate (wrong) because there is a constant level of offset in the system (the process of calibration will remove the offset and make it accurate and well as precise).


I must disagree. I’m not trying to be rude here, but it was you who brought those terms into the discussion.
I hope your spelling of tolerance and calibration isn’t deliberate.

_________________
Views expressed are personal opinions and are not necessarily shared by the Safe Speed campaign

Actually smeggy precision IS resolution in terms of mechanical instruments. The resolution of the measurement is a definition of its precision - ie a micrometer will have a resolution (smallest measureable quanitity) of 0.001 +/-0.0005mm. Therefore it is PRECISE to 0.0005mm. Its zero or basepoint is set by calibration, and if it is not calibrated correctly then it will suffer 'precise yet inacurate results'. The resolution will not chance but the correctness of the result will.

Like I said previously I think that we're looking at the same terms from different sides. I spent 3 years at Uni studying (amongst other things) instrumentation and control systems in mechanical engineering applications, and that's where I got my definitions from. However, if you're talking about signal generation rather than measurement then it is possible that the terms have different meanings.

Oh and BTW - I'm an engineer, not an english teacher!

_________________
Science won over religion when they started installing lightning rods on churches.

Ditto!


I still disagree. Some mechanical instruments are capable of being influenced, a speedo is a good example (EDIT: a micrometer can suffer from mechanical wear resulting with mechanical error beyond its resolution). An instrument can have a level of precision below that of its resolution due to noise/instability/drift/susceptibility. If a system can be influenced, the resolution won’t change but the precision can suffer.


This brings me full circle – the speedo in question wasn’t accurate, but it was (or may be) precise; ‘accurate’ was not the correct term to use in that instance.




‘Function over form’, being a fellow engineer I can certainly agree with that

_________________
Views expressed are personal opinions and are not necessarily shared by the Safe Speed campaign

But ACCURATE is still the correct word to use when describing the error in speedo reading in this case. All bike speedos have the same precision - +/-0.5mph. You cannot use the word PRECISE to describe error!

_________________
Science won over religion when they started installing lightning rods on churches.

I wasn’t using the term ‘precise’ to describe an error; I wasn’t describing any error (or anything negative) at all. I said ‘precise’ due to the “5% all the way through the range” statement.
The speedo is inaccurate because it has an error (off-center), but balrog used a positive term for the description (inaccurate didn’t make sense given the context), hence my comment.

_________________
Views expressed are personal opinions and are not necessarily shared by the Safe Speed campaign

But I still don't agree that the word precise is correct in this context. Maybe consistant or reliable, but not precise.

I don't think we're ever going to agree on this, so let's leave it eh?

_________________
Science won over religion when they started installing lightning rods on churches.

"precision: The ability of a measurement to be consistently reproduced. "

http://dictionary.reference.com/search?q=precision





Agreed that's something I really could do with right now!

_________________
Views expressed are personal opinions and are not necessarily shared by the Safe Speed campaign