WeRelate:Suggestions/Move Source Quality selection to Source page

Wondering if the source quality selection (i.e. primary, secondary, questionable, etc.) should be moved to the source page itself vs. the person page? That way it would only have to be entered once at that location vs. repeatedly for each citation. If there were any disagreements about quality, those could be hashed out on the talk page of the source page, which seems a more appropriate location. Thanks. --Cos1776 19:02, 24 September 2011 (EDT)

But some sources are both. Death certificates, for example, have both "primary" information (the death info) and "secondary" information (any birth or parents info). -- Amy (Ajcrow) 19:10, 24 September 2011 (EDT)

I think it should simply be removed altogether. Sources are rarely all one or all another. For example, the Newton genealogy transcribes lots of primary wills, but as a whole is secondary (also full of "I supposes"). VR Marlborough are full of entries based on Andrew Henshaw Ward's records (VR Nantucket similar use of William C Folger). Being primary doesn't rule out errors in one fact here or there, and I could dig up plenty of vital records that are wrong. Besides which, a common error is misapplying a correct primary record to the wrong person. I don't think applying some blanket rating is going to be of any help. Every application of a source has to be analyzed in conjunction with all the other evidence pertaining to that same fact, and in the end, the rating will not matter. --Jrich 20:07, 24 September 2011 (EDT)

These are all good points. I can see that a blanket designation is not always accurate, yet I do think it is worthwhile to encourage users to evaluate and disclose the sources of their information. Perhaps a better way of looking at it is that the quality designation belongs to the cited information for that individual and not to the source as a whole. If I think about it that way, I see that it does belong with the citation. -- So... I suppose that what I would really like to see is the ability to attach the same source citation to multiple pages, similar to the way in which you can attach a source record to multiple individuals in ancestry. It would cut down on a tremendous amount of redundant entry work and there could be a check box for quality designation for each individual (if you wanted to do that). Has this been previously discussed?

(ps - Please see also my request for the upload of an additional extension for selected transclusion. I am not sure why it was not user/day/time stamped in the suggestion table.) Thanks. --Cos1776 12:15, 25 September 2011 (EDT)

I like the idea of attaching the same source citation to multiple pages. To me, this seems like something I would do outside of editing any one individual page - like adding "Source citation" as an option on the "Add" menu, where you could enter the information about the citation and the titles of the pages you wanted to attach it to.

What about having it be a bookmarklet or a browser extension, like Evernote's Web Clipper? Would be helpful to have a bookmarklet that would let you copy text from a page on the web and let you create a source citation with that text and add it to one or more pages at WeRelate.--Dallan 14:33, 26 September 2011 (EDT)

I requested the ability to attach a common Source citation to multiple pages when I first joined WeRelate. At that point it was sort of shouted-down because I was a "newbie" and did not understand what I was doing (?). So rather than go through the work of entering the identical citations for half a dozen to a dozen immediate family members, I entered them sparingly. Sorry but I do not do GEDCOM uploads so everything I enter is done directly in the web pages. I still think this is a good idea and I would still like to see it happen. One example of my principle that it is better to design an interface that encourages and rewards users for doing the "right" thing than to provide reams of instructions/advice/criticism for doing the "right" vs "wrong" thing.--Jhamstra 17:36, 6 February 2014 (UTC)

I'm happy to see a discussion of source quality designation, since I've had problems with the present designations from the beginning. Specifically, a Source is not primary or secondary, a Source is original or derivative. It is the information in a Source that is primary or secondary. To use Amy's example of a Death Certificate, that is (presumably) an original source (the first record) -- unless it is a transcription, extraction, index entry, which are all derivative sources. As Amy points out, the information in an original source can be both primary (provided by an informant who was present or involved) and secondary (provided by an informant who was not present or involved), depending on the specific event. I suppose you could describe the information in a derivative Source as reliable or questionable, depending on the sources cited in that source, but I don't know that this has been discussed anywhere, given the emphasis on using original sources. (This is discussed in Elizabeth Shown Mill's book, Evidence Explained. I'm away from home, so cannot give page numbers, unfortunately.)

This means that description of a Source would not change -- it would always be either original or derivative -- so that description could go on the actual source page, but I would hate to lose some reminder to contributors of the need to evaluate their sources. I also think it would be nice if we used the correct terminology.--GayelKnott 10:21, 29 September 2011 (EDT)

I have a real problem with rating sources, period. Certainly one must be aware of one's sources and analyze their reliability, and this should be done by every researcher for every source, and certainly every time something is copied there is a slight increase in the chance of error, but I think any rating system, especially in a community environment, is potentially an over-simplification/misleading (implying that you don't have to do a thorough analysis because the rating says this is the highest quality source available) and source of pointless arguing (given the various levels of knowledge, and differing personal practices that everybody has adopted for their own research, any hope of a consistent application of rating criteria is a pipe dream).

Access to original records is difficult or impossible in many cases, and when I am looking for sources, practically I must often hope people are doing valid transcriptions or abstracts, or I will be severely constrained due to limited access. Also, I am aware that, for things like published vital records, the compiler is probably more competent to read the original than I am, and more likely to get it right. So, I tend to trust transcripts and abstracts as if original, until I have evidence they are wrong. Note that this is exactly the same criteria I apply to original sources. Even original records are occasionally wrong and blind acceptance of something because it is original is also a mistake. I have transcribed some vital records, and I know that sometimes copies are all that is available, and that sometimes copies and originals are intermixed with no indication. In some of these cases, it can take an expert to know which you have. WeRelate users are not experts (in the general case).

So, specifically, my sense is that your "derivative" rating is too broad, lumping useful and useless sources into one overly idealistic category. Trying to build in some rating system that can account for the complexities one runs into, which after all, make a difference in some very small percent of cases, as opposed to just relying on free-form discussions on the target page or its Talk, seems to me a waste of time. Especially when the weighted average of WeRelate pages is probably still unsourced in any form.

What I believe is the important distinction is whether it is identified what is the basis for a fact versus unexplained assertions. (Of course, by basis I mean some contemporary document like will, vital records, baptismal record, gravestone, etc., not merely passing the buck to a different source containing nothing more than unexplained assertions.) Once one knows the basis, one has the tools to follow the chain of evidence as far back as one feels is productive. The value of a cooperative community like WeRelate is that maybe there will one or two researchers who will invest the effort to do the proverbial exhaustive search on a page and post conflicting evidence if found. Then, source provenance and reference to the originals may be required to resolve the conflict. But for the benefit of future readers, who may have seen only one of the sources, and be unaware of the conflict, the process that led to the result probably needs to be summarized by a note or discussion, and not simply by rating this source a 1 and that source a 2. --Jrich 13:17, 29 September 2011 (EDT)

"Derivative" is not my rating, it's the professional standard. Analysis of the quality of a source is the first step in analysis.--GayelKnott 14:42, 7 October 2011 (EDT)

So is primary and secondary, as a rating of information quality, which tends to be orders of magnitude more important than original versus derivative in assessing the reliability of the data, based on my experience. And the rate of errors in original documents is not exactly zero, either, so in the end, the only effective way to maximize the reliability of one's data is the exhaustive search called for by the GPS. --Jrich 15:35, 7 October 2011 (EDT)

I'm not going to comment on the utility of rating sources, but as a historical note, the reason that source citations can have quality ratings at WeRelate is because source citations can have quality ratings in the GEDCOM standard. We just copied the GEDCOM standard. But removing the quality field has been suggested before.--Dallan 00:02, 30 September 2011 (EDT)

Although I did zero in on the source ratings initially, my interest is more in the mechanics and efficiency of the citation process,so I too will respectfully step back from this part of the discussion and continue on a different page.--Cos1776 15:48, 30 September 2011 (EDT)

I'm not a big fan of the "Quality" field because it seems too subjective. It's in there only because it's in the gedcom standard. But as far as I can tell it's not used much. What if we were to get rid of the Quality field in source citations altogether?--Dallan 16:57, 10 June 2012 (EDT)

Works for me. --Jrich 22:28, 10 June 2012 (EDT)

I prefer to see the Quality field stay. There are generally two cases where I use it. First is where I actually have or can inspect the original source document to determine its quality. In this case I will sometimes apply a Primary designation as appropriate. The other is like the first, but where there are obvious problems with the source document or citation. In this case I will flag it as Questionable with an explanation about why it is wrong or suspect. These two cases help others who follow my trail to understand my analysis - especially warning them of errors and pitfalls in various public records and published information.--Jhamstra 17:28, 6 February 2014 (UTC)

Dallan's word "subjective", plus the very widespread misunderstanding of what primary is (as opposed to original, which is something different) says to me it is pointless. And yes, some people like to mark sources questionable because they personally don't approve of them. I think this is inappropriate. One person I worked with once refused to use one source because they got his ancestor wrong, but in my case, they were right and the supposed expert was wrong, as we discovered after reading half a dozen wills. Even when a source is, in a particular case, questionable, a comment and explanation would be clearer, more educational, and allows for response, whereas checking a box is simply an pre-emptory assertion. --Jrich 18:24, 6 February 2014 (UTC)

I never mark a source questionable without an explanation (as I wrote above). For example my grandmother by her own account and those of others who were at the wedding, was married in New Jersey but the marriage was recorded in Michigan. The name of the town (in New Jersey) on the Michigan record does not even exist in that county in Michigan. Another example, is the recorded birth and death of one of my grandmother's brothers. They got the wrong name on the death record. The brother recorded on the death record was alive before his ostensible "birth" and after his ostensible "death" as shown by other public records, and also the grave markers. These errors were made in the days before actual birth and death certificates - only a county clerk listed births and deaths in a register - often a year or more after the actual events, in a rural area where there was no telephone and no way to verify the memory of the person doing the recording. In this case I suspect the original notes said something like "baby Wilson" and the clerk asked someone for the first name and switched two of the sons. I could give you many other examples but these should suffice to make the point. When one finds serious discrepancies in records I submit it is better to document them and flag them as problematic rather than to say it is merely "opinion" and give multiple "opinions" equal weight.--Jhamstra 19:33, 6 February 2014 (UTC)

I think we are in total agreement that even the best sources can have errors, and about documenting errors when found. I just question the value of this particular field after one has attached a big note and source references to a source citation explaining why you think it is wrong or questionable. I never mark a source at all, which is why I would prefer to see the abuses of this field done away with. I certainly don't think it is understood or used consistently when it is used. --Jrich 23:06, 6 February 2014 (UTC)

FOLLOW UP:
Source Quality has been removed as a Source Citation field.

This Suggestion will be moved to the archive, unless an objection is posted within 10 days. --cos1776 17:11, 27 October 2016 (UTC)