Talk:perfect gold standard test
Latest comment: 4 years ago by Kiwima in topic RFV discussion: December 2019
Wikipedia Edit History
edit* (cur) (last) 16:20, 21 February 2006 Stifle (Stub sorting - You can help!) [rollback] [proxycheck] * (cur) (last) 10:58, 16 February 2006 Manop m * (cur) (last) 12:53, 1 December 2005 Michael Hardy * (cur) (last) 12:36, 25 November 2005 171.66.222.155
The following discussion has been moved from Wiktionary:Requests for verification (permalink).
This discussion is no longer live and is left here as an archive. Please do not modify this conversation, but feel free to discuss its conclusions.
"(computing) A test that measures the value of a binary variable without error." In Google Books I don't find this, but do find a lot of medical usage, which would appear to be perfect + Gold standard (test). Equinox ◑ 06:45, 13 December 2019 (UTC)
- Measuring values is an activity in the realm of physics or other experimental sciences, not computing, so the definition does not make sense. That makes it impossible to verify it anyway; occurrences of the term cannot mean what it is claimed to mean. When the entry was created by dumping a Wikipedia stub not meeting the encyclopedic criteria for inclusion on the Wiktionary heap, it made no claim to being a computing term; the full content at the time was: “{{dicdef}} A perfect gold standard test is a test that, by definition, measures the value of a binary variable without error. {{Vocab-stub}}”. This makes it a concept in statistical testing. If we remove the inappropriate context label, the term meets the perfect gold standard test for being SOP, so what is there to verify? This should have gone straight to RfD. --Lambiam 09:58, 13 December 2019 (UTC)
- The difficulty is that however certain one might feel about the SoPitude of a term, it is possible that there is some evidence that it is not transparent in use. To collect the evidence more than a week (the minimum dwell time at RfD) is deemed necessary. In addition, RfD has tended to be citation-free. (One could say RfV is like a trial court and RfD like an appeals court or constitutional court.) One might argue that logical analysis of the definition is sufficient without any empirical evidence. This does not allow for the typically poor definition-writing skills of those writing their own definition. Gathering citations allows one to determine whether the term in use is actually not so transparent and then rewrite the definition to reflect that.
- I don't think that is a good way to look at it. That would require every challenged word to go through RFV before going to RFD, which is putting a lot of extra work on those of us who go looking for citations. The first definition had to be changed because it was a word-for-word copy of the definition in yourdictionary.com (plagiarism) (including the "(computing)". And yes, the quotes I could find (citations page)do not support restricting the term to computing literature, where I did not find any uses. But this seems a pretty straightforward RFV issue. Kiwima (talk) 18:51, 13 December 2019 (UTC)
- It is indeed sad that we have so many definitions that have no cites or only one old cite or reference to an old cite and yet so few spend time citing entries. DCDuring (talk) 21:12, 13 December 2019 (UTC)
- Suppose, hypothetically, that there is an entry for purple pebble, defined as: “A small stone with a colour that lies somewhere between red and blue”. Now it is entirely possible, as with all SOP entries, that somewhere out there there is a use of the term that is quite opaque – for instance, it means in the context, “An enchanted hallow that has great powers”. All well and good, but then it is not an attestation of this entry, but of some homonymous term. Attestations of the term purple pebble of the entry, used with the meaning “A pebble that has a purple colour”, will only underscore that this is a sum-of-parts. --Lambiam 23:12, 13 December 2019 (UTC)
- I think it is often just a question of how much overconfidence one has in one's understanding of the English language since 1470 and from the Phillipines to Guam (the long way). DCDuring (talk) 00:33, 14 December 2019 (UTC)
- No, Lambiam has it right here. When something goes to RFV, we do not search for all possible citations of all possible meanings, we look for support for the meaning that is challenged. If that meaning is SOP, any other meaning is likely to miss the sieve and it is just a waste of everybody's time because the entry will be deleted as SOP. Kiwima (talk) 01:46, 14 December 2019 (UTC)
- I think it is often just a question of how much overconfidence one has in one's understanding of the English language since 1470 and from the Phillipines to Guam (the long way). DCDuring (talk) 00:33, 14 December 2019 (UTC)
- Suppose, hypothetically, that there is an entry for purple pebble, defined as: “A small stone with a colour that lies somewhere between red and blue”. Now it is entirely possible, as with all SOP entries, that somewhere out there there is a use of the term that is quite opaque – for instance, it means in the context, “An enchanted hallow that has great powers”. All well and good, but then it is not an attestation of this entry, but of some homonymous term. Attestations of the term purple pebble of the entry, used with the meaning “A pebble that has a purple colour”, will only underscore that this is a sum-of-parts. --Lambiam 23:12, 13 December 2019 (UTC)
- It is indeed sad that we have so many definitions that have no cites or only one old cite or reference to an old cite and yet so few spend time citing entries. DCDuring (talk) 21:12, 13 December 2019 (UTC)
- I don't think that is a good way to look at it. That would require every challenged word to go through RFV before going to RFD, which is putting a lot of extra work on those of us who go looking for citations. The first definition had to be changed because it was a word-for-word copy of the definition in yourdictionary.com (plagiarism) (including the "(computing)". And yes, the quotes I could find (citations page)do not support restricting the term to computing literature, where I did not find any uses. But this seems a pretty straightforward RFV issue. Kiwima (talk) 18:51, 13 December 2019 (UTC)
- The difficulty is that however certain one might feel about the SoPitude of a term, it is possible that there is some evidence that it is not transparent in use. To collect the evidence more than a week (the minimum dwell time at RfD) is deemed necessary. In addition, RfD has tended to be citation-free. (One could say RfV is like a trial court and RfD like an appeals court or constitutional court.) One might argue that logical analysis of the definition is sufficient without any empirical evidence. This does not allow for the typically poor definition-writing skills of those writing their own definition. Gathering citations allows one to determine whether the term in use is actually not so transparent and then rewrite the definition to reflect that.
RFV-resolved. Someone may want to nominate this for deletion as SOP. Kiwima (talk) 22:01, 21 December 2019 (UTC)