[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Accuracy vs. precision (was APD V3 #992

Hello Dave,

You are, obviously, right that accuracy and precision are both important
(and good to have)! Unfortunately, in real life we often have to make
choices from the options open to us.

I have no quarrel with your rifle example -- one CAN correct for windage
and elevation for a bull's-eye shot. But, if after sighting in the rifle
one heats the barrel, so that it is slightly deformed, all bets are off.

A similar situation exists with pH electrodes. Depending on an
electrode's characteristics, any mishandling of the electrode will
change these characteristics and have a different effect on its accuracy
and/or precision. Some changes will be easily correctable, others not at

A well known example is the "dead" electrode (or one in a "stand-by
mode") -- no matter what the solution it is placed in, it always shows
the same pH -- ultimate precision, but zero accuracy! -- Can't think of
a rifle analogy for that one! :-) 

A more frequent cause of error is electrode DRIFT, due to many causes.
Drift can be time or motion dependent, or apparently completely erratic.
It is very difficult/impossible to compensate for. But at least one can
notice drift. A drifting electrode should be replaced! (I guess, that
the rifle analogy here would be that each round has a different load.)

If one has an accurate, but imprecise instrument, then the solution is
to take many successive measurements and average them. One is likely to
get close to the "true" value. (Many modern instruments do this



> Date: Sun, 25 Apr 1999 21:28:34 -0700
> From: Dave <datruedave at home_com>
> Subject: re: pH measurement (accuracy or precision?)
> Just because something is accurate, does not mean it is precise!
> I say that the most
> misleading instrument is one that is accurate but not
> precise(provided you have a way of
> calibrating your readings).  You really need both accuracy and
> precision.  Liken it to
> shooting a gun(bad example this week, but easiest for me to
> explain).  If you shoot 5
> shots and they are all within an inch of each other, but 6 inches
> away from the bullseye,
> then it is precise, but not accurate.  That can be compensated
> for.  If, however, you
> shoot 5 shots and there are some that are 3 inches away from the
> bull, some dead center
> and one that is 3 inches away in the opposite direction, then it
> is more accurate(each is
> closer to the target), but less precise than your first 5 shots.
> Some of the shots are 6
> inches away from each other!   There is no way to compensate for
> this.  When you
> translate this back to test equipment, you see that precision is
> really more important,
> but accuracy(or some way to calibrate and zeroize your equipment)
> is also vital.  I've
> never had a pH meter, I don't know how they work, but if they can
> be adjusted at all, or
> if they have a way of calibrating them(as I've heard), then I
> would go for the more
> precise instrument because you can always compensate for
> consistent inaccurate
> measurements.
> >
> > Date: Fri, 23 Apr 1999 10:19:38 -0500
> > From: George Slusarczuk <yurko at warwick_net>
> > Subject: Re: pH measurement (was APD V3 #986
> >
> > Hello James,
> >
> > What you should be after is greater *accuracy* in pH
> measurements. The
> > precision will follow. Using the common definition,
> >
> > Accuracy = how close is your reading to the "true" value,
> >
> > Precision = how close to each other are the readings,
> >
> > it follows that if you have good accuracy, i.e. your reading is
> very
> > close to the true value, then, of necessity, the individual
> readings can
> > not be too far apart. The most misleading instrument is one
> that is
> > precise, but NOT accurate!
> > <snip>
> > Best,
> > George
> - --
> - -Dave (in San Diego)
> ~~~~~~~~~~~~~~~~~~~~~
> I'm sure you can figure out my e-mail address.