Re: [Phys-l] propagation of error (or not)

[John Denker <jsd@av8n.com>]

On 06/01/2009 01:55 PM, alex brown wrote:
> I require a book on lab\practical practises and techniques including 
> error propagation at under graduate level.  We used one at UCL 
> (London), but can't remember the name of it apart from it being very 
> thin...

I've never seen a good book on the subject ... and I've seen quite
a few absolutely horrendous ones.

Most of them I consider sophomoric, by which I mean lacking in breadth
of experience, lacking in depth of understanding, tending to mistake
the easy case for the general case, and blissfully ignorant of their
own limitations.

I recommend
 -- starting with a review of basic probability, without which no real
  understanding of uncertainty is possible
 -- then introducing "crank three times" in a single-variable situation
 -- then proceeding to "crank three times" in a multivariate situation
 -- then proceeding to a simple Monte Carlo

Yes, I omitted any mention of manual / algebraic / analytical propagation
of error from this list.  That was completely intentional.  Such techniques
may work in end-of-chapter problems,	 but almost never in real life.

For the next level of detail on all this, see
  http://www.av8n.com/physics/uncertainty.htm
especially
  http://www.av8n.com/physics/uncertainty.htm#sec-propagate

> Thanks for any advice and apologies if this is not the correct place.

This is probably as good a place as any.  Certainly there are people
on this list who know what they're talking about.

>  I do not leke to rely on the reviews on Amazon to make a decision 

That is good, healthy skepticism.