A last note; listen to Markus, he is usually right. Darn! 😤 On Fri, Aug 12, 2016 at 12:02 PM, John Erling Blad <[email protected]> wrote:
> Latest date for population isn't necessarily the preferred one, it can be > a predicted one for a short timespan. For example Statistics Norway provide > a 3 month expectation in addition to the one year stats. The one year stats > should be the preferred ones, the 3 month stats are kind of expected change > on last years stats. > > Main problem with the 3 month stats are that they usually can't be used > together with one-year stats, ie. they can't be normalized against the same > base. Absolute value would seem the same, but growt rate against a one-year > base would be wrong. It is a quite usual to do that error. > > A lot of stats "sounds similar" but isn't similar. It is a bit awkward. > Sometimes stats refer to international standards for how they should be > made, in those cases they can be compared. It is often described on a page > for metadata about the stats. An example is population in rural areas, > which many assume is the same in all countries. It is not. > > And while I'm on it; stats often describe a (possibly temporal) connection > or relation between two or more (types of) subjects, and it is not > something you should assign to one of the subject. If one part is a > concrete instance then it makes sense to add stats about the other types to > that item, like population for a municipality, but otherwise it could be > wrong. > > In general, setting the last added or most recent value to preferred is in > general wrong. > > And also, that something is not-preferred does not imply that it is > deprecated. And also note the difference between deprecated and deferred. > > On Thu, Aug 11, 2016 at 10:56 PM, Stas Malyshev <[email protected]> > wrote: > >> Hi! >> >> > I would argue that this is better done by using qualifiers (e.g. start >> > data, end data). If a statement on the population size would be set to >> > preferred, but isn't monitored for quite some time, it can be difficult >> > to see if the "preferred" statement is still accurate, whereas a >> > qualifier would give a better indication that that stament might need an >> > update. >> >> Right now this bot: >> https://www.wikidata.org/wiki/User:PreferentialBot >> watches statements like "population" that have multiple values with >> different time qualifiers but no current preference. >> >> What it doesn't currently do is to verify that the preferred one refers >> to the latest date. It probably shouldn't fix these cases (because there >> may be valid cause why the latest is not the best, e.g. some population >> estimates are more precise than others) but it can alert about it. This >> can be added if needed. >> >> -- >> Stas Malyshev >> [email protected] >> >> _______________________________________________ >> Wikidata mailing list >> [email protected] >> https://lists.wikimedia.org/mailman/listinfo/wikidata >> > >
_______________________________________________ Wikidata mailing list [email protected] https://lists.wikimedia.org/mailman/listinfo/wikidata
