Search functionality needs improvement, please
I am not happy that I cannot find via search for a baptism record for an individual where the record has already been indexed. And because the record has been indexed, I cannot attach the record to the person on my tree. This is illogical and should not happen.
Here is the record that has been indexed: https://www.familysearch.org/ark:/61903/3:1:33SQ-GTW9-HCM?i=275&wc=SFV4-YWG%3A44975801%2C44975802%2C49449701%2C49453101&cc=1468076
I search for Bertha Maria Downing, and no records appear. Yet it exists.
Answers
-
On your first link, if you mouse over the "Bertha Maria" line on the Image Index tab at the bottom, a piece-of-paper icon will appear to the left of Bertha's name. Click on that icon to bring up her index detail page: https://www.familysearch.org/ark:/61903/1:1:QVS4-X1J6. On that page, click the "Attach to Family Tree" button at the right to open Source Linker, where you can attach the index entry as a source to Bertha's profile.
The reason your search is unsuccessful is that Bertha Maria has no surname in the index. Try entering only her given names, and put Downing in for her father's surname: https://www.familysearch.org/search/record/results?count=20&q.fatherSurname=Downing&q.givenName=Bertha%20Maria&q.givenName.exact=on. This search has one result, which is her baptism.
For future reference, you can attach images as sources even if they have index entries associated with them, regardless of any omissions or other errors in that index. You'd use the "Source Box" button on the image browser window (like your first link) to create the citation, and then you'd use your Source Box to attach that citation to all of the relevant profiles.
3 -
I wonder how many indexed records I am missing in my searches because I am constantly forgetting the need to make at least two types of search. Firstly, the "obvious" one - searching on the first name / last name of the prime person. But, secondly, if this doesn't produce the anticipated result(s), then you need to search on just the first name(s) of the prime person and input the last name under a parent's name - usually the father's, but against the mother if you know / suspect the child might be illegitimate.
I assume this advice is contained somewhere in a Help article, but it is so easy to forget if you are accustomed to finding records using the more obvious method of searching on the full name of the individual in question.
This "anomaly" comes about through the indexing process. I believe some project instructions advise the names must be recorded exactly as written. So, for example, "John Smith, son of William & Mary" presents no problem, but if the entry is recorded in the format "John, son of William & Mary Smith" some (but apparently not all) sets of instructions will insist the indexer records only "John" as the first name of the prime person, with the last name being indexed just against (father) William.
In my experience, this practice is relatively unusual, but must be taken into account, nevertheless. Even more confusing is the situation whereby baptisms of children to the same parents, in the time period and parish, can often be recorded in these two different ways. Hence, you might find some of the children by adopting a straightforward search, but have to change your search method to find the others!
Personally, I think common sense should be applied in these cases (i.e. indexing on what is obviously the last name of the child, as well as the father), but the general position with FamilySearch is to index "exactly as written". I have been told (by indexers) that their concern is only to record what they see and not to worry about any difficulties this might cause further along - specifically, once these records appear online and researchers attempt to find them.
1 -
@Paul W, the problem is that indexing considers so many different factors that the search algorithm's refusal to show "Bertha Maria" with parents named "Downing" as a match to "Bertha Maria Downing" gets shoved way down in priority.
Birth records -- both civil and ecclesiastic -- very, very often do not have a surname field for the child. They only have a column or field for the given name(s) of the child, with a different column/field (or two) for the names of the parents. Most of the time, it's possible to make a pretty good guess at what surname the child used based on that, but sometimes, it isn't. For example, a question recently came up on the Hungarian genealogy email list: the child's status of "illegitimate" was crossed out, and a man's name was added to the parents column; the Remarks section identified that man as the mother's "legal husband", but added another man's name as a person that the mother "had relations" with. So which of the three surnames should be attached to the child? It is FamilySearch Indexing's policy that such judgements are not the indexer's job, but the researcher's.
People being people, when an indexing project is set up with a child's surname field, but the records do not have such a field, indexers very often ignore all instructions to the contrary and fill in that indexing field based on their assumptions, resulting in the mess that you see: some entries have a surname, some don't, some have more than one (if the indexer couldn't decide). All of the newer projects I've seen therefore do not even have an indexing field for the child's surname on birth records.
The fact that the search can't find Bertha Maria Downing is not a fault of the index: the search algorithm ought to be able to match the surname field to all surname fields in the entry. I don't know why it doesn't. Perhaps the Indexing department needs to communicate with the engineers: Search needs to be fixed so that these kinds of records do not require gymnastics, neither from indexers nor from researchers.
3 -
@Julia Szent-Györgyi Thank you for your reply. I did search for Berth Maria without a surname and nothing came up. I suppose I should have also included her father's surname, but the record should have come up without the father's information.
I will investigate other methods to attach records that have been indexed as you suggest. Although I find the 'Source box' button difficult to use as it is not guaranteed that the relevant record at the right place within the record archive, gets attached to a person. The other method to click on a subtle icon next to the person's name in the Image Index box to attach a record is more successful. But the icon is so subtle and not exactly obvious that clicking on it will lead you to attaching the record to a person.
The point is that I, and other users, such as @Paul W , are not satisfied with the UX in searching for records. The search functionality changed I think late last year, but I don't think it is an improvement. I understand that the indexers play an important role here and the search function relies heavily on the indexed information, but I feel there are a few records that are not being picked up, as in my example I was struggling with.
@Julia Szent-Györgyi Are you part of the Family Search Team? After working in Family Search building my tree for several years, I do have some suggestions that I would like to make to help improve the site. I find it difficult to find where in the website I can make these suggestions, so that the dev team can consider them for their development roadmap and perhaps put on their backlog. Are you perhaps able to point me to where these suggestions can be posted for consideration?
Thank you.
0 -
@JohnBennett11, no, I'm not connected with FS -- I'm just a fellow user of their resources, like you and Paul.
Suggestions have a section here in Communities: "Suggest an Idea". We're told that engineers or product managers look at posts there, although we almost never see any evidence of that.
2 -
@Julia Szent-Györgyi Thanks very much for letting me know and for your help. I appreciate it.
0 -
If the newer projects do not have an indexing field for the child's surname on birth records, there could be a real problem. Some regions have unusual naming conventions.
Westphalia used the Hof naming system. Sometimes the child was given a name that was different from both parents. I can understand indexers making a mistake when the child's surname is not stated on the form. But if the child's surname is listed, the indexer needs to be able to enter the correct name.
0 -
@Cheryl Viering, I'm sure that indexing projects for documents that sometimes do record a child's surname do have an indexing field for it. This is one reason that indexing projects take a while to be set up: they have to carefully consider all of the possible permutations of what's in the documents and what indexers will do with outliers. As I said, all of the newer projects that I've seen do not have a child's surname field. I haven't seen all that many different projects. (I don't index in German -- I have a hard enough time with That Dratted Handwriting for my own research, never mind inflicting my attempts on the world at large.)
1 -
And just to add another permutation to the mix - I've found many records in my research where the child has a surname, but the father does not. That can lead to erroneous assumptions, on certain marriage records, if the bride is a widow/divorcée with a different surname from her father.
0
