Why can't I find a record that I know exists? Why is search so inaccurate and variable?
I'm trying to sort out my ancestor family Willis in Cambridgeshire England in the 1700s.
I need to find all Willis children baptised to Thomas Willis and Mary in Cambridgeshire. HOW?
One of the families on FamilySearch has a child called Mary, baptised 1755 in Linton Cambridgeshire, within a family where the parents were married in Linton, yet all of the other children were baptised in Stapleford, before and after this Mary. I think this couple is wrongly linked to all of the children except this Mary, so I'm trying to find others baptised in Linton, and confirm my family from Stapleford. This is the known record:
https://www.familysearch.org/ark:/61903/1:1:NR3W-31M?lang=en
You'd think that I'd find her and all of the other children of Thomas and Mary by search with these parameters WHICH MATCH THE RECORD:-
- last name: Willis (exact)
- Birth range: 1700 - 1799
- Father's first name: Thomas (exact)
- Mother's first name: Mary (exact)
- Country or location: England
- State or Province: Cambridgeshire
- Record Type: Birth, Baptism and Christening.
- https://www.familysearch.org/en/search/record/results?count=100&f.recordType=0&q.birthLikeDate.from=1700&q.birthLikeDate.to=1799&q.fatherGivenName=Thomas&q.fatherGivenName.exact=on&q.motherGivenName=mary&q.motherGivenName.exact=on&q.recordCountry=England&q.recordSubcountry=England%2CCambridgeshire&q.surname=willis&q.surname.exact=on
IT DOES NOT. It finds 15.
If I untick the exact from the two parents, it finds 1468 records. If I then at the top select Birthplace: UK and Ireland it reduces to 20. They all have Thomas AND/OR Mary OR none as parents. One of the extra records clearly matches the exact criteria but didn't show with exact ticked. It still does not find this KNOWN record of Mary.
I've also tried just searching for Mary Willis without parents, with just a year range, and then narrowing down using the births filter at the top. Only one of two known records is found.
THE RECORD THAT I KNOW EXISTS IS NOT FOUND BY SEARCH. Another version (either the Parish or the Bishops Transcript) of it is, but that does not have the village name because of abysmal transcription where the whole film has NONE of the towns transcribed, even though they exist at the start of every section.
How SHOULD I be searching to find all children of Thomas and Mary Willis in Cambridgeshire?
Answers
-
Not all records survived to be microfilmed and then digitized. Some survived but are not legible, with water damage or faded pencil entries. Some may not have been indexed to enable searching by name. Standardized spelling is a very modern concept.
I started my search with the record set Parish registers for Linton, 1559-1876 https://www.familysearch.org/en/search/catalog/koha:209754
My search parameters were set to accommodate abbreviations and variants. Since you know that one child was born/baptized in 1755, I set my date parameters to 20 years each way. If Mary was the first child, the mother could conceivably have had a child as late as 1775. If she was the last child, as early as 1735.
I used Willis as the surname for the child, Tho* for the father, and Mar* for the mother. I set no names to exact. I have 8 results.
3 -
I'm glad there was a good solution for your situation @JamesBarton9.
However, I have sometimes been frustrated with the Search function, and changing parameters doesn't help. More than once I have found a record on Ancestry that I want to find on FamilySearch so I can use SourceLinker. I know exactly what the record looks like so I don't have to use wild cards. The Search function won't find it searching for one person in the record, but it will find it after searching for another. I don't know why Search sometimes fails this way. 🤔
0 -
Thanks Áine, I do understand the record limitations, but the record exists and is transcribed on Family Search. The issue is that search is not returning that record that matches the criteria of the search. I gave the link to the specific record. What I cannot do is find it by search.
If I search that Linton image set you suggest, it finds a different entry - the Parish Register. Worse than that, the individual record does not state Linton as the parish - merely giving the county. Even though it came from the record set for Linton. That is bad indexing, and EXCEEDINGLY unhelpful. I only know it is Linton because of the other record. The collection information only shows England, Cambridgeshire, Parish Records. If I had not come across that record through your route, there is no way to know it is of that Parish.
This is the other record. The entry I should be able to find is apparently in the collection England, Births and Christenings, 1538-1975. This transcript of the record includes the Parish name.
BUT, if I search that collection https://www.familysearch.org/en/search/collection/1473014 using place as Linton, Cambridgeshire, England, it is NOT found! In fact, I cannot find that record at all , or indeed any for Linton, by searching that collection, whatever selection of search terms or options I use.
Searching that specific collection as you'd expect to do for Mary Willis born Linton, Cambridgeshire, England in 1755 only finds a 1753 baptism in West Wickham, and 24 records where Mary is the mother - but the dates aren't even close to 1755.
I am trying to find an overall search that will actually return details from the whole of Cambridgeshire, because there are Willis families all over and there are very common names and couples. Some have been wrongly merged in Family Search so the tree isn't reliable.
1 -
Hi Cheryl
Actually it hasn't - that only finds the record I can already find by search. See my later post.
Jim
0 -
I can understand your exasperation with this matter, as I have had similar problems in the past.
In your example, there appear to be a few separate issues:
(1) That no records for Mary Willis can be found when the placename Linton (Cambridgeshire) is specified, in spite of it showing in the source you illustrate with your screenshot.
(2) The "England, Cambridge, Parish Registers, 1538-1983" collection from which the 1755 baptism can be found does specify a parish in many cases, but - for some unknown reason - not in many others (where just "Cambridgeshire" is shown).
(3) Even the title of that collection is poor, as it implies the records are just for the city of Cambridge, by leaving off the suffix "shire". The "How to use this collection" link provided takes you to a Wiki page that does make it clear that the collection covers the county of Cambridgeshire and not just Cambridge itself.
However, there are similar problems (e.g., the 1851 census collection for England and Wales) in the example of just the county being shown in some cases, but the specific location being given in others. That problem has been around from when I started using FamilySearch and remains unresolved in spite of many collection updates and reports from users of the problem.
These matters should really be escalated for investigation / correction and, hopefully, a moderator will pass these problems on to the teams concerned.
1 -
I could not find any problem with how Search retrieved results within the parameters you applied. Here are some explanations for why you didn’t locate certain records, especially the one you knew existed.
- A padlock icon usually means the record isn’t searchable; while there are rare exceptions, this isn’t one. The indexed document is a type-written extraction from Linton parish records, created in 1981. Its 1981 dating likely explains the restriction.
- In the Stapleford parish records (7907898), pages skip from 1752 to 1755, so anyone born, married, or deceased during 1753–1754 won’t appear in searches. Although the Bishop’s Transcript for those years survived (7676702), that film hasn’t been indexed.
I can understand the problem of indexes listing only the county name when the images are only available at a FSC or AL. These parish records were all created/indexed in August of 2023 and that is how they were originally indexed—no changes have been made. However, the metadata for each film does specify Linton or Stapleford. I will bring this up with engineers.
I hope that answers your questions. Our Search engineers continuously look for ways to improve the algorithms but most often the issue is in the records or in overly specific criteria. When I search as Áine suggested, in a “smaller pond” with broad search terms, I feel more assured that I’ve found all there was to find.
1
