Talk:Experimental uncertainty analysis

Latest comment: 4 years ago by Luifer Hoyos in topic Figure 4. Increasing in standard deviation

First cut

edit

Still needs some work, esp wikilinks, jumps to the figures, etc but I wanted to get the bulk of this online. This is a far more accessible, IMHO, article on the subject than are (1) the one titled "Uncertainty analysis" (which, forgive me, says nothing much at all!) and (2) the one titled "Propagation of Uncertainty" (which is a collection of equations, with minimal discussion).

I will provide a detailed table of example eqns in a few days. No doubt in an article this long I've missed some things, and some of the material certainly can be improved. But I really think that it will be more useful than the existing articles to your average folks (especially students) who want to learn a bit about this.(Granted, these "average folks" need to know some calculus!) Rb88guy (talk) 15:12, 12 March 2009 (UTC)Reply

Rev 2

edit

Added one set of equations for the various calculations (for a univariate power function). Lots more to do on this, esp since I check the equations with simulations. There are certainly a few things that need some clarification, and I'll get to those as time permits. Rb88guy (talk) 18:17, 13 March 2009 (UTC)Reply

Rev 3

edit

Added a section showing the derivation of the PoE equations. I have not seen this in any text other than Meyer and even there it isn't real obvious what's going on. Also I have not seen anywhere but in Rohatgi the E[z] expression, which is needed to get to the Var[z]. The E[z] gives the bias, which seems like it might be useful. Next is to do a couple more univariate PoEs and then a few bivariate ones. Rb88guy (talk) 18:38, 16 March 2009 (UTC)Reply

Rev 4

edit

Done with univariate cases; not convinced that the expo and log ones are of much use, since the quality of the approx varies wildly- it can be excellent, or worthless. Next is bivar. Rb88guy (talk) 17:13, 17 March 2009 (UTC)Reply

Did two bivars, don't think I'll do any more of this unless there is a request. This should be about it aside from some clarifications here and there. Rb88guy (talk) 20:43, 17 March 2009 (UTC)Reply

Typo fixed

edit

Thanks to 130.245.221.74 for catching a small typo in one of the equations. Glad to see someone is looking at the article, and closely enough to find this...Rb88guy (talk) 14:45, 28 October 2009 (UTC)Reply

edit

added, that shows PDFs of estimated "g" for Method 1 vs. Method 2 data processing, as discussed in the article. Vary the scatter of the measurements and the sample size, see effect on bias. Rb88guy (talk) 01:05, 18 December 2009 (UTC) There must be some material missing from the beginning of Section 1.1 (perhaps an accidental delete?) There is reference to "the length of the pendulum" when no pendulum has been introduced, and reference to Eq(2) when no equation has been given, etc. 108.32.88.226 (talk) 23:43, 19 February 2012 (UTC)Reply

Gave evidence of missing material in Section 1.1

edit

Section 1.1 refers to the length of a pendulum when no pendulum has been introduced, and refers to Eq(2) when there is no Equation(2) (or Equation(1)). It appears that a recent edit deletd more than was intended.108.32.88.226 (talk) 05:27, 21 February 2012 (UTC)Reply

October project

edit

Banner infers textbook sounding. how about - para remover 'It will be seen that' since you discuss it. — Preceding unsigned comment added by Andys'edtits (talkcontribs) 13:17, 22 October 2013 (UTC)Reply

fix section - formulas not parsed

edit

The equations in "Example expansion: p = 2" inside the Derivation of propagation of error equations fail to parse. Could the original author please fix it?

Figure 4. Increasing in standard deviation

edit

In the last paragraph, it says: "...consider Figure 4, where the standard deviation of the time measurements is increased to 0.15s, or about ten percent." Shouldn't it really be 0.033 or something like that? Luifer Hoyos (talk) 22:57, 26 November 2019 (UTC)Reply