Thursday, October 25, 2012

Trying Out the RootsSearch Chrome Extension - Part 2

I posted Trying Out the RootsSearch Chrome Extension - Part 1 yesterday, showing how to download and activate the RootsSearch tool that can be used to search a number of websites for records of a person from within another website (e.g., while in FamilySearch, then search or Find-A-Grave).

Following up yesterday's post, I want to check out what happens when I click on the links in the RootsSearch extension.  Here are some of my search results for records of Lucretia Townsend (Smith) Seaver, one of my second great-grandparents:

1)  Here's the Search fields filled in for Lucretia Townsend Smith (1828-1884), born in Medfield, married to Isaac Seaver, with parents Alpheus Smith and Elizabeth Dill (this is where I ended in Post 1).  I consider these terms to be typical of what a researcher might enter when offered a search form of this nature:

2)  I clicked on the blue "Ancestry" button on the RootsSearch screen, and a new tab opened and I saw:

You can see the search terms transmitted by RootsSearch to in the upper left-hand corner.  The only thing different from what I entered was the death year - I entered 1884 and it used 1883 for some reason.  The search used the Ancestry Ranked Matches algorithm (non-exact) for all search terms.  There were over 8 million matches.  It found Lucretia's birth record, her 1850 US Census record, and her 1851 marriage record easily, but not her census records with her married name in 1860, 1870, and 1880, and not her death record in 1884.  I didn't check every match...but it looks like the search did not find Lucretia with her married name Seaver in any records but the Member Trees.

I could edit the search terms on the Ancestry screen - when I clicked on "Exact matches" for the search terms entered, it found only Ancestry Member Trees.  That is not unusual - actually it's to be expected, since I over-specified the search field information.

My lesson learned here is that a user really needs to search for both a maiden name and a married name for females.  Users should not over-specify search field entries, and be ready to edit their search fields to narrow their search.  The search for males is much more reliable and complete, unless the user over-specifies the search field entries.

2)  Back on the RootsSearch screen with the same Lucretia Townsend Smith search terms, I clicked on the blue "FamilySearch" button and saw:

You can see the search terms used on FamilySearch.  A plus/minus 10-year range was used on the birth date and death date.  The spouse's first name was used, but not the spouse's last name.  The parents names were added (not shown above).

This specific search found Lucretia's birth record and the 1850 US Census record.  The 1851 marriage record was far down the list, to Isaac "Leavir" rather than "Seaver."

Again, the lesson learned here is the same as above - females are difficult, don't over-specify, and edit the search terms in order to narrow your search.

3)  Back on the RootsSearch screen with the same Lucretia Townsend Smith search terms, I clicked on the blue "Find A Grave" button, and saw:

I clicked on the "Refine Search" link on the left, and changed the search terms to be FirstName="Lucretia" and LastName = "Seaver" and searched and received two matches, including Lucretia Townsend Smith Seaver:

I could have used a MiddleName = "Townsend" to get the match, but did not get a match with  FirstName = "Lucretia Townsend" and "MiddleName" = Smith;" or with FirstName = "Lucretia" and MiddleName = "Townsend Smith;" or with FirstName = "Lucretia" and MiddleName = "Smith."

Again, the user must be prepared to change the search terms to reflect a female's married surname, and to modify or delete a middle name or maiden name from the search.  Again, males work better.

4)  Back on the RootsSearch screen with the same Lucretia Townsend smith search terms, I clicked on the blue "WeRelate" button and saw:

All of the search terms I specified are in the search fields on, but the 1884 death date has been made "31 Dec 1883" and the dates have a range of plus/minus 5 years.

The first result on the match list is the correct Lucretia Townsend Smith who married Isaac Seaver.  

5)  I expect online family trees to be found by RootsSearch when the person's name, birth data, death data, spouse's name and parents names are specified.  I did not check if a match comes up, because I don't have a current subscription.  My guess is that it would, since Lucretia Townsend Smith is there with that name.

My conclusions from this search effort are:

*  RootsSearch works fairly well, but not perfectly, at this time.

*  The user must have subscriptions to the subscription sites that the tool accesses (, and at this time).

*  The search terms can be edited by the user in order to narrow a search.  If the user changes any search field entries, they need to click on the "Update Links" button.

*  Users should not over-specify search terms.  Start with first name, last name, birth year and birth state/country.

*  Each website searched receives the pre-entered search terms specified by the user (although some sites don't get all of the search entries entered).  These search field entries need to be carefully reviewed and edited if necessary to narrow the search.  The user can use Exact Search for one or some or all terms, or use wild cards for names, if desired.

*  Users should do searches for female maiden names, and their married names, in order to find all of the records that match the person.  Obviously, users should not expect matches for records of females after marriage, except for records that might list her maiden name in a child's birth, death or marriage record.

*  RootsSearch has real potential, but it needs the users to understand the limitations of the tool.

Thank you to Justin York for creating a useful genealogy search tool.  I hope that he continues to improve it, perhaps add more record collection websites, etc.

The URL for this post is:

Copyright (c) 2012, Randall J. Seaver

1 comment:

Unknown said...

Thanks for putting both parts 1 & 2in a simple manner for all to understand! Just printed both articles out and started a folder for this for future reference!