Talk:List of the verified oldest people/Archive 10

RfC: Original Research

Are the Gregorian calendar notes and sorting by "age in days" original research or not? I will post my opinion on the situation and then notify all involved users below this summary.

Requestor's statement

Quite simply, I believe the notes for Izumi and Butariu are entirely speculative and violate Wikipedia's no original research policy. First of all, we have to represent the sources faithfully - we cannot speculate what they have or have not taken into account. As I said above, Wikipedia's job is to report what it sees, not to think about it, find new things out about it or discover flaws or things that you think the sources might have gotten wrong. Everything on Wikipedia must be verfiable; there's no room for us to say "well maybe they forgot this" even if we acknowledge that they may not have. The notes themselves are unsure whether or not they should be there, for they contain the caveat "This assuming the date of birth has not been otherwise adjusted as no mention of this anomaly is made in the source". Just because the source doesn't list every last thing that they took into account doesn't mean that they didn't take this into account. Now it just so happens that the source itself, Robert Young, contacted me and backed me up on this source. While I do realize that it was probably a mistake to allow a banned user to have his say on Wikipedia policy (and I have corrected this mistake), in this case a little WP:IAR applies, particularly because he is not being used to influence Wikipedia decisions, but specifically as evidence that my claims about original research on this issue are valid. He email is reprinted here for convenience, and it was originally posted on Wikipedia with his permission:

I'm not convinced that Shigechiyo Izumi needs a 'Gregorian calendar' adjustment. Ages in Japan were recorded using their own calendar; the 'year' is based on the year of the emperor's reign. Thus, Kamato Hongo was born in the year M20 (20th year of the Meiji era) which is then transribed into the English year '1887'. Since the transcriptions would have been modern (application to Guinness in 1978 for Izumi), but the original data were recorded in a Japanese system, I don't see how the Gregorian calendar would affect this process. In any case, the research was done in 1978 (I was four years old) and I had nothing to do with it, and the data for Izumi is now 'lost.'

As for the age in days ranking, a very similar argument applies. We have to maintain the ranking provided by the sources or else we run the risk of violating WP:OR. Reposting an argument from above, we're taking the ranking numbers from a certain source, the GRG. We cannot deviate from their ranking without it being original research - sure it may seem logical with the number of days on the side, but if we allow for unstructured deviations, we're introducing our own element of original research and potentially subjectivity. If we begin to deviate outside the structure of the original source (and, indeed, since the methods we use to include people who are still living and move them up the list are within the original methodology, this is acceptable), then we open the door for people to add their own, more harmful original research to the list, such as adding the 150 year old claimant from Timbuktu. If the GRG ranks them at the same level, we have to respect the original sources; we can't just decide that they made a mistake and rank them differently. We've researched and found that some people lived through leap years and some didn't. Therefore, the wish is to modify the "official" (as defined by this page) ranking and deviate from the source. This is research is original. "Rank" is not our ranking, it's the GRG's and, unless they change it to account for tis research, then "Rank" has to reflect the GRG. Maybe it's not the best policy, but OR has to be followed, whether it's on a WP:BLP or something somewhat more trivial such as this, and there's no room for WP:IAR. Again, we have Robert Young, the source, giving us an example of exactly why we should be following the sources rather than speculating. The same caveats for the above email apply to this one as well:

Regarding the discussion of rankings based on supercentenarians (years and days):

1. The GRG chose the year/day format, in part due to tradition. This tradition includes Guinness World Records but was actually started by T.E. Young, president of the Society of Actuaries (London) in 1899.

2. In many cases we do not know the exact day/hour someone is born. Someone may appear to be older at '41,821' days than someone who is '41,820' days. But let's suppose that person A was born at 11PM and died at 1AM, they lived 41,820 days and 2 hours, they get credit as 41,821 'days'. But the second person was born at 1 AM and died at 11PM, they get credit for 41,820 days, but they actually lived 41,820 days and 22 hours. Thus, based on the scientific concept of 'sigificant digits', we shouldn't really be over-focusing on extreme/exact amounts like this...we simply don't have enough information to conclude who is older.

3. The 'equal' rankings were also done to conform to Microsoft Excel, which uses the years/days to calculate the ages.

I'd appreciate it if you could point this out, that it is not simply a matter of the GRG not knowing math.

Sincerely, Robert Young

Thus, my argument is that a) The notes on the Gregorian calendar should be removed as they are completely speculative and ORish and b) The "rankings" should remain faithful to the original sources, although a fair compromise is to include "age in days" as a column and let people decide for themselves when they come to the article. Cheers, CP 23:07, 3 May 2008 (UTC)

This seems to me to be a rather lengthy argument to a rather non-issue. I'll address the "age in days" column first. It's merely there as addtional information. It in no way deviates from the original source, being the GRG. The ranking follows "years and days" without regard to what appears in the "age in days" column. In that several individuals have noticed the small inconsistancies I do feel it is important for comparison. And as for original research, any grade school child with a calender can produce the same material.

As for the footnotes, I do think they provide necessary information where there may be a lack of clarity from the original source, being the GRG. The information provided is a compilation of different areas of information on Wikpedia condensed to provide a possible conclusion. I don't believe it's original research. It's provided so the reader has the best information sources to make an educated decision. That's all then. Thank you. TFBCT1 (talk) 04:16, 4 May 2008 (UTC)

I agree with Robert. Extremely sexy (talk) 15:30, 4 May 2008 (UTC)

It seems to be a non issue if you ask me. It's just there for additional information. I wouldn't really call it original research. --Npnunda (talk) 02:30, 5 May 2008 (UTC)

That's also very true. Extremely sexy (talk) 22:03, 5 May 2008 (UTC)

So you are of the opinion that when the originator of a source that Wikipedia cites contacts a Wikipedia administrator and claims that we are misrepresenting their source, that is a non-issue? Cheers, CP 04:17, 6 May 2008 (UTC)

Some countries did not adopt the Gregorian calendar. Because of this, Peoples ages may be different by a few days. This footnote has nothing to do with the grg not knowing math which is what the email states. It is my humble opinion that the footnote would be there even if we used a different source. This is just my "good faith" opinion and I'm not an admin

like you but that's how I feel about it. Regards --Npnunda (talk) 04:12, 8 May 2008 (UTC)

I wouldn't call it original research. The GRG supplies the date of birth and the date of death. The length between those dates can realistically only have one answer: We just need to take into account any calendar changes. The issue of time will always be a factor however they are ranked. If we rank them by years and days then someone who died on their birthday, technically 6 hours before the time they were born would be listed as living another year. I think we should use the GRG as a key source, but I don't think we should be copying everything from there. As has been shown with leap years, some of the information isn't in the best format. I disagree with point 2 because as I have already mentioned, time will always be a minor factor however you list people. However, ranking by just days has a greater resolution. As for point 1, it's more informative for the public to see it listed as years and days, as they will obviously not know how long 44000 days is off the top of their heads. However, from a scientific point of view, we should be using the most accurate measurements to rank them. Practices have to be changed if you want to make progress.SiameseTurtle (talk) 11:28, 11 May 2008 (UTC)

It doesn't matter how easy or obvious it is to work out someone's age in days, if it's not quoted from a source outside wikipedia then it must qualify as original research. If the GRG or another reputable source has a list by days then it could be included here without further discussion. While including a column for days may provide some clarification it may also be a distraction and perhaps may even be confusing for some readers. Personally, I would not be concerned if the "age in days" column were removed until such time as it can be quoted from a legitimate source.DerbyCountyinNZ (talk) 00:49, 12 May 2008 (UTC)

OUTSIDE OPINION There's a guideline mixup occurring here. This isn't a case of OR, it's synthesis. The guideline on that deals with improper synthesis. It's in no way harmful to do math. If we have a source that says "there were two boys and two girls in the classroom at the time", it would be accurate to enter at WP "there were four children in the classroom". Unless the source is using trinary numbers, the math is accurate. JeremyMcCracken (talk) (contribs) 05:09, 30 May 2008 (UTC)

I have to agree with the thrust of this argument, that day-counts, even if accurate, constitute original research. Indeed, since there is a discrepancy in some cases between people as counted by year/day and just days, the ordinal ranks are slightly altered if one goes by days. And people coming onto this page might wonder why Bettie Chatmon is ranked above Odie Matthews when they both lived the same number of days. While there may be a valid argument to go by days instead of years/days, that is not what GRG, the source of these lists, goes by. We should, therefore, omit the day column and if some don't like that, then appeal to GRG to alter how they rank ages. Because as it is we have two methods side by side which don't match. Canada Jack (talk) 14:54, 3 June 2008 (UTC)

Further to what I have said, this is in my opinion original research for the simple reason there are different ways of making a count and the source here has opted on one count. Day counts was not the method used, even if there is a good argument to make that this is a more accurate way to do it. The appropriate thing here to do would be to omit the day count and make the case for a change to GRG. When they change their count method, then we can do the same, but not before.

It is not our duty to list what we or what some here believe to be the more accurate count method, it is our duty to accurately reflect what those who have done the research have concluded. And they, like it or not, go by the year/day count. I do agree that it is appropriate, however, to make note of potential discrepancies in terms of Julian/Gregorian usage. But even there, GRG simply has noted how many days after the claimed birthday the person lived which is accurate in and of itself, though the actual day count may be out by 11 days or so. "But that person is older/younger than claimed!" some might say. Well, yes and no. If one goes by year/day, no, by day count, yes. The problem here is a "year" is not a fixed, standard length, but a day is. But GRG uses years. It's not up to us to "fix" it, however. Canada Jack (talk) 16:39, 6 June 2008 (UTC)

I reverted what someone had decided to do - change the rank to list these individuals by day count, not year/day count, at least until this issue is resolved. But are people here aware that GRG in fact does have a day count list, which differs from the one we have here? [1] If you look at his in Explorer, it reads as a year/day list. But in Firefox the listing is wonky with overlays of day counts, and many of these day counts are fractional. Additionally, the day numbers don't match what is on our page. The year/day counts do match (though the list is as of Oct 2006). Canada Jack (talk) 18:42, 6 June 2008 (UTC)

why bias toward US citizens ?

Given that the us is ~~ 5 % of the world population, and certianly no more then ~ 1/3 of the "developed worlds" population, it seems odd that more then 50% of the entries are listed as us. Am I just miscounting ? Is this better health in the US in the late 1800s or that WW1 and WW2 were not fought on our soil ?Cinnamon colbert (talk) 01:37, 17 May 2008 (UTC)

It's a combination of good records, good health, and large population. Don't forget that these also need to extend back 120+ years. Countries that are developed now weren't necessarily developed 100 years ago. The USA was, as was the UK, France etc. but they had smaller populations. WW1 and WW2 were probably factors too, especially for men. Imagine how many men in the UK would have reached Henry Allingham's age if they hadn't died young.SiameseTurtle (talk) 11:49, 17 May 2008 (UTC)

Does the rank update itself?

I am aware that the age of the living people automatically updates itself but when they overpass a deceased person in the ranking, does it automatically rank them one rank further up, or must you edit it to do so? --90.206.20.104 (talk) 16:27, 1 June 2008 (UTC)

Someone has to edit it to do so. Luckily, we have many diligent editors watching this page. I can't even remember the last time I had a chance to update it! Cheers, CP 18:40, 4 June 2008 (UTC)

List Totals

Is there any real point to this section? The numbers of living/deceased and male/female are already mentioned in the header. And what is the point of adding up the countries? How exactly can you have 0.5 of a country? Given the changing status of various territories, states and countries and the differences for birth and death how can the same weight be given to each? It all seems pointless and unnecessary. DerbyCountyinNZ (talk) 23:08, 9 June 2008 (UTC)

Unless there are any objections I'll go ahead and remove this section tomorrow (2 weeks without comment would indicate disinterest in whether it should be kept or not). DerbyCountyinNZ (talk) 00:09, 22 June 2008 (UTC)

Nelle Hunt

Nelle Hunt is listed as joint 99th with Helen Stetter both died age 113 years, 195 days. However, Hunt has lived for 41,468 days where as Stetter has only lived for 41,467. This would mean they both died at the same age but Hunt lived 1 day longer. Wouldnt this make Hunt 99th and Stetter 100th? I think this should be the case. After the recent addition of Catherine Hagel, shouldnt Stetter be eliminated from the list? --90.206.20.81 (talk) 19:35, 10 June 2008 (UTC)

GRG lists these people by year/day, and ranks them accordingly. While there may be a good argument to make that counted by days, Stetter should be off the list, this is not how GRG sees it. So, to rank them that way would be original research. And, to further complicate matters, GRG has done a calculation of days for these people (from 2006) and the counts differ from what appears on this page. For this reason I believe we should eliminate the day counts until GRG decides otherwise. Canada Jack (talk) 21:00, 10 June 2008 (UTC)

I am agreeing with Canada Jack. I initially thought it made sense to posibbly remove Helen Stetter and just keep Nelle Hunt in the 100 position, but our agreement with the GRG is to view individuals by their "year and days" totals irregardless of what appears in the "age in days" column. So Helen Stetter and Nelle Hunt are looked at as being equal. So until someone else appears on the list, both will remain tied in the 100 position. Thanks. TFBCT1 (talk) 23:18, 10 June 2008 (UTC)

Duplicate ranks

I've re-done the filling in of duplicate ranks, which was reverted earlier along with the "by days" ranking. I defer to the balance of arguments on the ranking point, but believe that duplicate ranks need to be filled in as they now stand. Sorting does not work properly with rowspanned rankings, beause the rowspan gets applied after the sort. As an example, on the rowspanned version, sorting by "Birth" results in garbage. 86.4.200.143 (talk) 20:39, 10 June 2008 (UTC)

Sorry about that. I knew there might be formatting issues involved with the revert, but the more important thing was to restore the original ranking until that issue can be resolved. Hopefully, fixing the boxes to allow sorting wasn't too complicated. Canada Jack (talk) 21:04, 10 June 2008 (UTC)

Please do not remove rowspans in lieu of multiple entries with the same ranking. This is not needed and is very distateful. It makes the page appear sloppy and amateurish. And I'm not sure who this is aiding. I habitually make all changes in rank to the table. It is done with ease typically around 8:15pm EDT the prior evening. Anyone else have an opinion? TFBCT1 (talk) 23:10, 10 June 2008 (UTC)

Perhaps someone here can steer me correctly - when I hit the sort buttons on the various columns, I get rather useless ordered ranks - births starting alphabetically by month with April, then August - not sure what this is needed for. If those columns are to be relevant should they not have information presented in such a way as to make relevant rankings? Like dates in a format like yyyy-mm-dd, names by surname, etc? Canada Jack (talk) 15:55, 11 June 2008 (UTC)

We've had lots of discussions on the sortable table, and I think that the general consensus was that it was unnecessary. Archive 5 has the most concentrated discussion, but there's a few comments scattered all over. Cheers, CP 01:33, 13 June 2008 (UTC)

New record

Walter Breuning has recently outlived Emma Taylor. Please update both US longevity articles to reflect this. 208.76.245.162 (talk) 07:37, 30 June 2008 (UTC)

This page lists by birth, and Walter Breuning was born in Minnesota. Therefore Emma Taylor is still the oldest born in Montana. (Yubiquitoyama (talk) 15:16, 30 June 2008 (UTC))

Do you remember?

Robert Young is in the process of validating a woman who then would make it a literal top 100 again by removing both women in 100th place currently, but I don't remember her name nor exact lifespan: who can fill in the gaps for me? Extremely sexy (talk) 14:24, 1 July 2008 (UTC)

Could this one be it? http://health.groups.yahoo.com/group/Worlds_Oldest_People/message/10141 SiameseTurtle (talk) 16:15, 3 July 2008 (UTC)

Yes, that's her. Extremely sexy (talk) 13:24, 4 July 2008 (UTC)

Any progress with this? TFBCT1 (talk) 01:41, 9 July 2008 (UTC)

Well: I will have to ask Robert, but he is busy with his thesis. Extremely sexy (talk) 12:24, 9 July 2008 (UTC)