Some people eat, sleep and chew gum, I do genealogy and write...

Thursday, October 5, 2017

Searches and Searches and Searches! Oh My!

Lions, Tigers, Bears - The Wizard Of Oz - YouTube

When I am asked about my genealogical specialty, I generally respond, "I find things." After all, that is the whole idea of doing historical research, i.e. genealogy. The last two days, I spent considerable time simply explaining to people how to find things; mostly their ancestors. One particular benefit of having huge online database programs is their ability to search through the content of those databases. One of the inquiries I received was a question about which websites would be best for doing research in the United States? In response, I tried unsuccessfully to elicit further information about which part of the United States the inquirer was interested in. But ultimately, the answer was the large online database programs.

In thinking about all of the questions I received, I realized that it had been some time since I had reviewed the large database search engines in a blog post. I thought it might be interesting to go through a sample search in each of the four large online programs;,,, and As I have done in the past, the idea is to take a somewhat random name from my family tree and compare the results of simple searches between the four programs. To make this fair, I really need to do more than one search with more than one name. So, here we go.

I decided to focus on searches in the United States and England. All four of the large online database programs have substantial records in the United States but because of, I decided to add in a second search in England. I further decided to choose someone in the mid-1800s that way I could be assured that there were potentially a lot of records available. I also decided to accompany the searches with some of my usual commentary. One factor that made selecting individuals to search was the issue that I have already researched nearly all of my ancestors in that time. I also decided to eliminate anyone who already had dozens of sources attached to their entry in the Family Tree. I also wanted to choose someone who had relatively complete basic information so the results of the search would not be overly affected by a lack of information. Finding someone with only a name and the date and a vague place takes more than a good search engine.

 For my first search, I decided to choose David Sheppard (b. 1802, d. 1877). Here is his detailed summary from the Family Tree. I need to note, that there were no sources attached to this person in the Family Tree.

To be fair, I have to note that this person already had three Research Hints from But the idea here was to see what happened if someone came into the program without any other information and simply did a search. But I do need to include the Record Hints in the number of responses. Using the Family Tree makes this rather simple because there are links to the four websites that use the information in the family tree to find additional records. Further, because I happen to have an LDS account with, I have access to all four programs. 

Once I got started, I decided that I would only consider fairly obvious sources. In other words, I would not comb through pages and pages of results. I also decided that I would not add any additional information for this initial comparison. I would also not change the default settings in any of the search parameters available in each of the programs. Essentially, I wanted this to be as close as possible to the perspective of a search done by a relatively unsophisticated user.

So here are the results:
394 results with six immediately identifiable sources.
There were only three immediately identifiable sources out of 387,320 results. However, upon clicking on one of the sources, identified five additional applicable sources.
There were two results in both of these were identifiable sources. As a note, there were both references to US census records.
There were 5,646,916 results. All of the initial results were from MyHeritage Family Trees. Since the rules of the game here mandated that I do nothing more, at this level, I would have rejected all of the results. However, recognizing the way that the program works, information that would not be readily available to a novice, I filter the information to look for census records and found a record. However, to be completely fair I would have needed to evaluate the information already contained in the large number of user family trees matches that were found.

The results from this are not obvious. All four of these programs are structured to provide automated record hints. But since my idea here was to show what happens if you do not have a family tree in each of the four programs and in addition, do not have advanced search skills.

Now let's see what happens with the person from England. I selected William Parkinson, (b. 1828, d. 1892). Here is a screenshot of his information from the Family Tree.

I would have to note, that the birthday, as recorded, is undoubtedly wrong. It cannot be the same as the christening date. I should also note that there were already seven sources attached to this person on the Family Tree.  There is also a little twist here and that this person died in Australia. But, in the interests of proceeding from the vantage point of a complete novice, here's the search.
On an initial search with 227 results, only one of the seven existing sources was found. As a matter of comment, none of the results included any records from Australia. Because I am familiar with the program, I realized that I could increase the number of accurate responses by adding in information not included in the original search. But to do this, I would have to have considerable experience using the program.
The initial search returned 197,587 results. Two of these were the search subject. However, both of the records were from Australia. In this case, clicking on the results did not produce any additional records.
An initial search only returned two results in either of these were the target person. Again, in all fairness, I already know that I found many additional valuable sources about this family on
Once again, I got a huge number of initial results, 1,148,828, most of which, initially, were MyHeritage family trees. There were two results to the Family Tree. Because there were sources attached to this individual in the Family Tree, these connections would have been useful. Additionally, examining the other family tree connections may have produced more information supported by sources.

In conclusion, it would seem that an absolute novice with some basic information about the person who lived in the mid-1800s would have a very good chance of finding pertinent sources to support existing information and add information. But it became abundantly obvious that real progress could only be made with some understanding of each of the four programs.

It would appear that each of the programs does an adequate job of finding records based on a name, a birth date, and a place but expanding on that information requires some degree of sophistication and using each of the programs. From the results of this test I would conclude that all four of the programs are very useful I would also suggest that having a family tree in each program would measurably increase the utility of each of them.

No comments:

Post a Comment