Wikipedia talk:About valid routine calculations

This essay started with contents transferred from some consolidated discussion of the WP:OR. --Krauss (talk) 21:45, 14 July 2013 (UTC)

Objective

The aim of the essay is to centralize (consolidate) all rules and recommendations about valid "Routine Calculations" and valid "summary of numerical data".

Correction suggestions

(remember that you can colaborate directly editing page, it is still a draft --Krauss (talk) 00:36, 18 August 2013 (UTC))

Round and precision

• `source say "5300 meters" and wikipedist write "17,388 ft"' — probably incorrect! "17,388 ft" usually means something like "17,388±0.001 ft", while "5300 meters" often means something like "5300±100 meters". Boris Tsirelson (talk) 18:10, 22 July 2013 (UTC)

Indeed. Wikipedia should not create illusory precision. Given, a figure of 5,300 m, "17,400 ft" is, in my opinion, right, and "17,388 ft" is dubious. Of course, given Wikipedia's worldwide use, a figure in feet should never be given on its own, except where feet are an international standard.

If a source says "120, out of a total of 200," then "60%" is in my view quite OK (but note comment on WP:SYNTH by Arthur Rubin below, which I agree with). -- 202.124.74.5 (talk) 07:15, 24 July 2013 (UTC)

• thanks Boris Tsirelson and 202.124.74.5: I edited, see if now ok! I also added a "Round and precision" section. --Krauss (talk) 01:18, 18 August 2013 (UTC)
• Exactly 2 inches is the same as exactly 5.08 centimeters; but to equate "about 2 iches" with "about 5.08 centimeters" is stupid nonsense. The latter tacitly claims to be accurate to the nearest hundredth of a centimeter. Michael Hardy (talk) 17:20, 18 August 2013 (UTC)

General comments

• "Well-knowed" should be "Well-known"

• "17,4 ft" should be "17,388 ft"
• And it should be added that calculation is only valid if all the data comes from the same source; e.g., if one source says 120 are positive and another source says there are a total of 200, then 60% is synthesis.

— Arthur Rubin (talk) 08:56, 23 July 2013 (UTC)

Thanks Arthur Rubin, I edited, see if now ok. I also added a "Mix sources" rule (5). --Krauss (talk) 01:18, 18 August 2013 (UTC)

• That essay has some issues, and I don't think it in any way accurately reflects the many discussions that WP:CALC has been subject to. For instance, it is typically not valid to do units conversion without preserving significant digits. The example in the essay is a source that says "5300 meters" with the conversion to "17,388 ft" (while it arguably should be "17,000 ft"). Also, averaging data is not always appropriate when the data is drawn from different sources (there was a discussion about this somewhere), since different data collection methodologies make taking an average a novel synthesis. These are just the things that jump out at me. Sławomir Biały (talk) 10:18, 25 July 2013 (UTC) at this old talk.

  I added here, at talk page, #Objectives for new people understand... The corrections and the new "Round and precision" section was commented above. About "averaging": I added (for consensus discussion) a "rule 5" about "valid mix" of data from different sources. About "discussion (...) somewhere": it is important to list here links to all points/discussions, to consolidate all here (see #Objectives). --Krauss (talk) 01:32, 18 August 2013 (UTC)

• I don't know much about this policy debate on Wikipedia. But with respect to graphical representation of data, there is an example figure of a histogram. Choice of bin size and starting point can dramatically affect the visual appearance of a histogram, and hence its interpretation by the reader. Judgment is certainly involved. I think that the data would have to be already binned in the original source, for depiction of such a histogram to be regarded as a routine calculation. Mgnbar (talk) 03:11, 18 August 2013 (UTC)

• Also, the verbiage about recursive combinations of routine calculations should be made less broad. Arguably any computation by a digital computer can be regarded as a complicated combination of elementary arithmetic operations. Mgnbar (talk) 03:11, 18 August 2013 (UTC)

• I agree this has serious issues. Where I can follow it I don't see it agreeing with existing guidelines or previous discussions. But mostly it's too difficult to follow: it's unclear what it's recommending or the point(s) it's trying to get across. Clarity is the most important aspect of any essay or guideline, whether or not it's complete, whether or not it summarises existing practices or tries to establish new ones. Otherwise when referenced or referred it can make discussions harder to resolve as different editors interpret it as supporting different sides of the debate.--JohnBlackburne^words_deeds 11:38, 18 August 2013 (UTC)

• Working definitions and examples gives rules of thumb that are completely off the table once any dispute comes up about any edit employing a Valid Routine Calculation, because they're not positions that editors can claim applies or judge for applicability, hence they are not adaptable and the cases where they may be reasonable don't make their extension to universals reasonable. WP:NOTOR#Simple calculations is formatted so as to be explicit about what isn't meant to happen, and leave making sure it doesn't up to the interpretations of editors. Working definitions and examples defeats that purpose by imposing an interpretation. This interpretation's suitability depends on the context an author is used to consulting synthesis guidelines in, and can be better summed up as a list of things to take for granted as obvious. But obvious has a history of being used for things that are only obvious to the author, which not only might be wrong, but with the author not providing the means for others to tell, could only be dismissed by default from others failing to verify it. So obvious calculations are properly part of the bold/revert/discuss strategy for getting things done. That also means if something's truly obvious, it won't need to be said what makes it obvious or condoned explicitly. In the reverse, certain data might be downright non-obvious but nonetheless have an obvious way of verifying it. A large part of elucidating technical articles is in fact pointing out derivations experts find elementary to non-experts, falling under "so trivial it couldn't be published" kind of routine calculations. This is what the guidelines for routine calculation mean in this context, which is not at all what it means in the context the article is aiming for, which could be incomparable to any number of contextual adaptations the guideline has seen. These are also contexts in which unbounded recursion of anything to obtain an unreferenced result is going to constitute original research. Furthermore, the criteria for including more than elementary arithmetic operations and unit conversions is artificial, and it's vague about whether it can extend to problematic situations, like formally valid statistical (re)representations that can be used to misrepresent data and the source's conclusions. If the data isn't presentable verbatim, only conclusions made by the source of the data should be made in the article, as interpreting data is exactly researching that data. Part of statistical research is the actual argument that data should be processed a certain way and that the resulting model of the data is a valid interpretation of it. ᛭ LokiClock (talk) 16:08, 22 August 2013 (UTC)

Consensus and voting, first round

The first section is an introduction for definitions, the second, Recommendations, is the major contribuition of the essay, and need consensus:

• Routine calculations
  • General rules:
    1. Added calc: please comment on the above and also indicate Support or Oppose in your comments.
       • ... your vote...
    2. Exact calc: please comment on the above and also indicate Support or Oppose in your comments.
       • ... your vote...
    3. Synthetic calc: please comment on the above and also indicate Support or Oppose in your comments.
       • ... your vote...
    4. Prefer calc: please comment on the above and also indicate Support or Oppose in your comments.
       • ... your vote...
    5. Mix sources: please comment on the above and also indicate Support or Oppose in your comments.
       • ... your vote...
  • Customary justifications: (edit the article for enhance or) please comment on the above and also indicate Support or Oppose in your comments.
    • ... your vote...

• Round, precision and other treatment procedures: under construction.

• Summary of numerical data: under construction.

Title nit pick

I would drop "About" from the title. People already know that the title tells you what it's about.--RDBury (talk) 04:06, 18 August 2013 (UTC)

Mistake

"Source 1[1]: 1500, Source 2[2]: 1600, Source 3[3]": 2000 is "translated" to "experimental values average at 1700 with a standard deviation of 0.9 [1-3]" — No, the standard deviation surely is much bigger than 0.9. Boris Tsirelson (talk) 08:39, 20 August 2013 (UTC)

Grammar and spelling

Basic grammar and spelling errors:

• "wikipedist" should be "Wikipedia editor"
• "A routine calculation consist of" should be "A routine calculation consists of"
• "well-knowed" should be "well known"
• "convertion" should be "conversion"
• "Any of these formulas must applied into" - missing word ?
• "the work not stops here" should be "the work does not end here"
• "When source have complex or technical data" should be "When sources have complex or technical data"
• "some numerical treatment can be make" should be "some numerical treatment can be made"
• "with an add your "translation" near to it" should be "with an added "translation" near to it"
• "If most if not all cases" should be "In most if not all cases"
• "the relative weights of each value is needed" should be "the relative weight of each value is needed"

Gandalf61 (talk) 14:25, 20 August 2013 (UTC)

Well, the original author isn't a native English speaker, so feel free. ᛭ LokiClock (talk) 14:02, 22 August 2013 (UTC)

You're welcome. Yaris678 (talk · contribs) asked for "input and suggestions" so I have pointed out the most obvious errors. I am not going to try to rewrite this essay so that it makes sense - that would be a much bigger job. Gandalf61 (talk) 14:17, 22 August 2013 (UTC)

Oh look I just fixed all those things you mentioned in 5 minutes or so. You're welcome. ᛭ LokiClock (talk) 10:22, 23 August 2013 (UTC)

Well, that's very kind of you, and I hope the original author appreciates what you have done. My own view is that when something is in as bad a shape as this essay is, then the original author should make some effort to sort out their own mess, rather than expecting other editors to do it all for them. Gandalf61 (talk) 10:49, 23 August 2013 (UTC)

trigonometric identity

Would a short elementary proof of a trigonometric identity qualify as a "valid routine calculation"? Michael Hardy (talk) 22:50, 22 August 2013 (UTC)

For example:

${\begin{aligned}\sec(\alpha +\beta +\gamma )&={\frac {1}{\cos(\alpha +\beta +\gamma )}}\\[10pt]&={\frac {1}{\cos \alpha \cos \beta \cos \gamma -\cos \alpha \sin \beta \sin \gamma -\sin \alpha \cos \beta \sin \gamma -\sin \alpha \sin \beta \cos \gamma }}\\[10pt]&={\frac {\sec \alpha \sec \beta \sec \gamma }{1-\tan \alpha \tan \beta -\tan \alpha \tan \gamma -\tan \beta \tan \gamma }}.\end{aligned}}$

If that can't be found in print, is it fair game? Michael Hardy (talk) 22:56, 22 August 2013 (UTC)

Featured article examples

The "Wikipedia tradition of use" or a "Wikipedia's well accepted form" criteria, need sampling to accomplish consensus. ~25000 Wikipedia articles was elected as Good article or Featured article, so there are many good samples. --Krauss (talk) 20:37, 8 September 2013 (UTC)

Starting with featured ones:

• Time data, not a exact year: “Taoiseach Seán Lemass considered reviving the Order during the 1960s”, (1.1)

• Romans: “(...) as well as the date of the Order's foundation in Roman numerals ("MDCCLXXXIII")”, (1.2)

• Metric convertion and rounding: “(...) this was perhaps due to the their size, 30x36 cm (12x14 in)”, (1.3).

• counting: “The robes of The 4th Baron Clonbrock, the 122nd Knight of the Order”, (1.2)

• interval, percentage and unit convertion: “(...) occurs at a rate of 1,000 to 1,000,000 molecular lesions per cell per day.[1] While this constitutes only 0.000165% of the human genome's approximately 6 billion bases (3 billion base pairs)”, (2.1)

• interval definition: “absorbed from blue/UV light (300–500 nm wavelength)”, (2.2)

• data of sources by range: “Five different studies found that between 40% and 90% of colorectal cancers have reduced (...) .[69][70][71][72][73]”, (2.3)

• "exact" copy of data (replaced "or more" to "about"): “This happened about 20,000 years ago,[230]”,(3.1). Extracted from source 230's abstract: “effectively has a single origin, which we dated at 20 000 years ago or more”.

• "double around", by data and by reference's differences: “Amphibians first appeared around 364 million years ago,(...) and modern humans around 250,000 years ago.[272][273][274]”, (3.2)

(aleatory selection of) Featured article's fragments used in the quotations:

1. Order of St. Patrick or permalink
2. DNA repair or permalink
3. Evolution or permalink
4. ...

About what you can add

See Extensive and intensive properties, indicating magnitudes where the addition make sense, and where not...

Even summing "cows with cows and fotons with fotons" (ex. "2cows+3cows=5cows" make sense, but "2cows+3fotons" not), the sum is valid only in the extensive magnitudes. You can sum cows, fotons, volumes and areas, but not temperatures or densities. The "extensive/intensive concept" supply some general rules about the valid use of the Routine Calculations.

--Krauss (talk) 05:07, 15 September 2013 (UTC)

About what "wikipedia calculators", and "choose of the correct template" or "correct use of each template function", see Help:Calculation and wikimedia.org/Help:Calculation.

Dispute resolution question for 2014 Israel–Gaza conflict

There is a disagreement on the 2014 Israel–Gaza conflict about whether an editor can assume casualty figures for civilians and combatants when the original source does not give them and there is still the possibility that they are unidentified. If someone here has some experience with these types of issues your help would be greatly appreciated. Thank you.Monopoly31121993 (talk) 08:15, 10 August 2014 (UTC)