Showing posts with label ancestry.com searches. Show all posts
Showing posts with label ancestry.com searches. Show all posts

17 May 2014

Are the Search Results at Ancestry.com Based on Phonics? Maybe....

A recent blog post on Ancestry.com's search result sort order ("Search Result Not in Alphabetical Order") prompted Sue Maxwell to comment that the results were actually sorted phonetically. That made perfect sense because in the initial post the last name of "Phair" was listed in the "F" section so that searchers of "Fair" would find it.

That made sense and I thought I had it figured out until I went to find another image to illustrate a follow up blog post.

Why is the last name of Perry listed in between Penfield and Penell? It would seem reasonable that Penfield and Penell would be adjacent as they "sound similar." Certainly Perry sounds different.


I understand why "Saint Andrew and St Clare" are listed together. But the other names notated on this image confuse me.

Which leaves me confused about the sorting of the search results. Strict alphabetical order I can handle. But now when I need to leave the surname blank on a search, I will have to scroll through all the results. Just in case.

I don't believe that we should take "whatever results" we get and be happy with them. There's no reason why these search results cannot be displayed in a way that makes sense. This is computer programming where sort procedures can easily be applied.


Ancestry.com Search Result Not in Alphabetical Order?

I'm confused again. [partial followup here]

I'm trying to find Sarah and Margarat Demar/Demarrah/Desmarais in the 1905 New York State Census on Ancestry.com. The sisters should be in somewhere in Clinton County, New York. That's not why I am confused.

Given the exponential number of spellings for the last name, I decided to perform first name searches in my attempt to locate the girls. The screen shot shown below was obtained when searching for individuals with the first name of Mar* born in 1888 (plus or minus 1 year) in Clinton County. It's not the search results that I'm confused about.

It's the way they are displayed that is confusing. I thought the listings were in alphabetical order by last name, but that does not appear to be the case. At least not this time. Mary Depre and Mary Deome are out of alphabetical order and Mary Phair is out of order as well (although admittedly "Phiar" is phonetically equivalent to "Fair" and may appear in the "F" section for that reason.) Even if one concedes that "ph" and "f" are equivalent, the names are still not in strict alphabetical order.

Why does it matter?

Because sometimes when performing first name only searches, I get more hit than I can ever review. But I will go to a section of the list that starts with a specific letter and manually review those searches. This is because Ancestry.com won't allow me to perform wildcard searches on last names with only one letter--so I use the alphabetical sort to essentially serve the same purpose.

But now I'm wondering just how alphabetical the sort is.

Now that I review the image above, I realize that Mary Chrystal is out of order also.

What am I missing? Or can Ancestry.com  not sort alphabetically?

And, as usual...the concern is: "if these are weird things that I notice, what other weird things are there with search that I am not noticing?"

[search results accurate as of 16 May 2014]

23 April 2014

The More Specific I Am the More Hits I Get

Again, I am confused by Ancestry.com's search. 

Ok, I'm not really confused. I know it's not working the way it is supposed to and I'm irritated. I love the wildcard options we have, but there's simply something about the search algorithm (or at least the error message) that is incorrect. 

It was fairly simple. I wanted to search for Johann(es) Grass who is believed to have made the trip from Germany to the United States in the 1880s. I'm not exactly certain of the year, but the Ostfriesen native appears as a sponsor for a grandchild in Illinois in the 1880s. Those details formed the basis of my search. 

Not certain of what port Johann would have used for his arrival, I chose to search all of the "Immigration & Travel" records at Ancestry.com by searching as shown below.


There were only 132 hits. This is not so many that I cannot go through all of them, but I decided that since Johann was around 1815, I would refine my search to concentrate on immigrants of that age.

And so I added 1815 (plus or minus ten years) to my search.That should give me fewer hits and make it easier to concentrate on those items that match my more narrow search.

Except I got "Too many matches."

How can there be too many matches? I've refined my search even more than it was originally constructed.

This search is more narrow than the one that resulted in 132 hits. I should have obtained fewer than 132 hits.

How can less than 132  be "too many?"

This is not the first time I have received results of this type.

I don't want to be told to "refine" my search--that's what I did.

I don't want to be told to search specific databases--after all, we're told now that "broad" and "global" searches are the way to go. That's how searchers get a more "positive user experience." I have a "positive user experience" when search responds in a way that makes sense.

There is no reason why this search should return "too many hits."

My problem with this is two-fold:

  • There simply are not too many hits to display them.
  • If Ancestry.com cannot get this search to work the way that it should, are there other searches (ones that don't give errors) that might not be working correctly? It makes me wonder if I'm missing results (and maybe there were "other" Bernards on my search from yesterday that simply didn't show up). My conclusion that I had "my" Bernard yesterday was partially based on what search results were returned to me by the search at Ancestry.com
Many of us make conclusions based upon search results at various sites. We may not want to admit it, but we do....especially when we are looking for a needle in a haystack and don't have any other contextual names upon which to draw conclusions. That's the case with Bernard--he traveled alone. 

I just wish Ancestry.com could make this search work the way it should work. 

I don't need smart phone apps. I don't need the ability to research at 3 a.m. on my cell phone in my bed when I'm half asleep. I don't need to know where my DNA was 2,000 years ago.

And I don't need lectures on Boolean searching--which is what I got from a staffer a few years ago. I have a master's degree in mathematics--I understand Boolean logic, ands, inclusive ors, exclusive ors, nested searches, order of operations, etc. etc. 

I just need this damned search to work. 


27 March 2014

A Leaf is Found 100 Years After It Falls from the Tree

I know some people love them, but this is why I am not an overly big fan of the "leaves" at Ancestry.com.

Results like this are also why I am not a fan of any of Ancestry.com's "broad" based searches. There is casting a wide net and there is attempting to drape the entire ocean in fishnet.

The results as shown below for Ekke Flesner border on the absurd. A man who in my database dies in 1833 is a match for a naturalization in 1880 and New York Passenger list in 1950?

I realize that I may have the death date of Ekke Flesner incorrect, but even if I'm off 100 years that still doesn't explain the 1953 passenger list.
"Leaf" results on Ancestry.com for Ekke Andreesen Flesner died 1 Aug 1833--results obtained on 27 March 2014.
The parameters on the years of the record and the year of the death must be set pretty far apart. I realize some people are happy for any match they get. I also realize that crazy matches like these frustrate new researchers. 

As for me, I prefer results that are reasonable. If Ancestry.com expects me to wade through the leaves, they need to do some pruning of their search parameters. 


09 March 2014

Broader Searches Give Me Fewer Results with the Ancestry.com Slider

If I'm doing something wrong, please inform me.

The broader a search is, the more results one should obtain. That's a basic premise of database searching.

If you want more results, make your more broad.

That's not what is happening for me at least on this search with the sliders at Ancestry.com

This search screen shows that obtained 194 results for my search of the 1900 census for John Coffmann as shown below--the last name is set to the second slider from the left.


I then broadened my search by moving the slider to the left.

Except I got 155 matches. Shouldn't there be more?



Am I doing something wrong?

Can anyone else replicate these searches?

I'd like to know what I'm not doing correct or what I am not understanding correctly.