Path to Online Image: Waypoints vs Search Terms

Looking at two examples in QuickSheet: Citing Ancestry Databases & Images:

For the "City Directories: Images" example, waypoints to the image are given in the citation.

For the "Draft Registrations: Images" example, search terms are given to find the record/image ("imaged card for Clovis Julian, no. 120, New Orleans Draft Board 13").  I looked this up at Ancestry and found the waypoints to this image to be Louisiana > New Orleans City > 13 > Draft Card J > image 393 of 425.

Is there any preference or rationale of choosing to list the waypoints or to instead list the search terms to find the item of interest?  



Dear EE,

I've read QuickLesson 25 *many* times :)

1.  I'm still having trouble understanding the difference between the "World War I Draft Registration Cards" and, say, the "California Passenger and Crew Lists" example.  Upon navigating to these collections, both provide search fields and results that bring the researcher directly to the image of interest.  But the "California" example adds the waypoints.

I *can* understand the difference pointed out further in the Lesson "The type of database has changed. It's no longer 'database with images.' Now it's 'browsable images.' This term is used for a set of records with no indexing and no neatly typed abstract or extract of the record. We have to browse the images to find what we need."  (bold added)

2.  I also have a difficult time understanding why the Draft Registration Cards citation isn't positioned as a layer 2, with the actual draft card cited as a layer 1.


Submitted byEEon Fri, 02/12/2021 - 11:31

Jeff, your questions focus upon two issues

  • Whether to cite the search term or path > waypoints
  • Whether to assign the document or the database to Layer 1, with the other appearing in layer 2.

As QuickLesson 25 states under “Bottom Line”: technology has made it so much easier to access the records that are our lifeblood. At the same time, it has also complicated the act of citing sources.

Search term vs. path

There is no rigid rule as to which we must use. We choose whichever one best fits the situation.  Most users choose whichever is the simplest for the particular data set.  Sometimes it simpler (and much shorter) to cite the search term. Sometimes, especially when the search term would generate many hits, it's more logical to cite the path.

Arrangement of Layers

There is no rigid rule as to which we cite first—the database that offers the record, or the record in which we are interested. We choose whichever one best fits the situation.

Citing the database in Layer 1 is typically done when

  • We are using that database for numerous documents, and we prefer to have just one Source List Entry, rather than creating a Source List Entry for each document; or
  • The provider’s image set presents documents or pages without adequate identification so that we do not have enough detail to create a full citation for the document. Instead, the only identification we have of where that document came from is whatever the provider tells us. (Yes, on many occasions, the provider’s details do err.*)

Citing the image in Layer 1 is usually done when

  • We want the document or the record book or the file to be a standalone item in our Source List; and
  • The image is part of a set that has been imaged completely enough that we can identify for ourselves what it is we are using.

I could stop here and say:  EE and the QuickSheets demonstrate both approaches.

However, you ask for the logic for specific examples, so let’s take the three examples you cite and analyze the logic in each:

WWI Draft Registration Card:

Consideration 1: We can use the search term “Clovis Julian” to find the document. Or we could browse. The database offers both options, but to locate Julian by browsing, we would we would have to go card-by-card through 13 different districts in New Orleans, not knowing which district he is in—and then we’d find him in the last one. The logical first approach, when a browsable set of images offers a search option, is to search for the name.

Whether we use the search engine to find Julian’s card or use a path, once we have arrived at the card we are looking at one card in isolation. We can’t flip back to the start of the image set and see the title of a register, or the ID of a file and the collection it's in, or the name/number of a published set of microfilm.  If we go back to image 1 of that image set, all we see is another card. We have no idea what that card is part of or who created them or where they are found or what collection they are a part of.

How do we, then, construct a citation for the card? To do that, we would have to “borrow” Ancestry’s citation. “Borrowing” citations—i.e., silently copying whatever source data is claimed by the source we’re actually using—is not a sound research practice. In some situations, it is unethical. In all situations it is unwise.  (On this point, you might type “borrow” into the search box on this page and read further on the issue.*)

Consideration 2: Most researchers using this set of images will be using it for numerous individuals. Creating a separate Source List Entry for each individual card could significantly bloat the Source List. Creating one master source for the database and then citing the specific card in the “item of interest” field is logical.  (In the “old days,” when all work was done manually, we could handle this issue any way we choose. In today’s world, where millions of researchers create citations within their own database or within a citation manager, we have to choose which item we want to feature.)

City Directories:

Again, whether we cite the imaged item in Layer 1 or cite the database in Layer 1 is a matter of choice.  This image set is also accessible via browsing or a search term, but the structure of the database differs from the WWI Draft Registration database in one significant regard.

Whether we browse or use a search term, we arrive at an imaged page. At that point, as we eyeball what it is imaged, it is easy to identify what we have, in order to create a citation for Layer 1.  The entire publication is filmed. We can flip back to the start of the book and get all the citation data needed for a full citation in Layer 1. The simplest approach is to cite the book in Layer 1 and cite the website in Layer 2.

Alien Crew List:

Again, this image set is also accessible via browsing or a search term, but the structure of the database differs from both of those above.

Whether we browse or use a search term, we arrive at an imaged sheet. As we eyeball what it is imaged, it is easy to identify what we have, in order to create a citation for Layer 1. We can scan back a few pages and forward a few pages and see the entire document.  We know we are using a record that is titled “Alien Crew List,”  for the S. S. Arrino, which debarked at Lota, Chile, and arrived at San Francisco on 7 October 1913. We see all the pages and know that our C. W. Dendy is on p. 11. 

We can flip back to the first images in that dataset and see that we are using images from microfilm. We have a microfilm number so that we can verify what the provider is citing. We have a letter of certification from Immigration and Naturalization Service telling us exactly what we are using, &c.

In short, from eyeballing the images, we have enough detail that we can construct a citation to the ship roll itself, to create Layer 1. However, the provider also gives us source details that we cannot verify from the images that we see. The provider tells us that the National Archives/INS microfilm M1416 has a title Crew Lists of Vessels Arriving at San Franciso and that the provider’s images were actually made from FamilySearch’s microfilm copy of M1416, which FamilySearch identifies as microfilm 4482913.

Because the provider gives extra details about the source that we cannot eyeball for ourselves, we add a Layer 3 to say that the provider is “citing …..”

Again, whether we choose to feature the ship roll in Layer 1—citing each ship roll we use as an individual, standalone source—or whether we choose to cite the database and put individual ship data in our “item of interest field” is our choice.

Lumping vs. splitting

This choice has long been described by researchers as lumping vs. splitting.  Some researchers prefer to “lump” all entries from a database into one master citation in the Source List. Some researchers prefer to “split apart” all the individual records they take from a database and cite each one separately in their master source. 

EE and the QuickSheets demonstrate both approaches.


*As an aside: most users of EE would be surprised at how many times, in the preparation of this guidebook, when I backtracked a provider’s citation to the original (as I did for each of the 1100+ examples), I found the provider’s citation pointing to an entirely different source. Humans create those citations for the provider's database, and humans err.

In each case, of course, I contacted the provider, pointed out the error, and asked if it would correct the problem before EE went to press so that EE would not have to explain their error.  Obviously, the errors were quickly corrected. Just as obviously, we have to ask: are we willing to "borrow" someone else's citation without verifying that the citation is correct?

Submitted byJeffH13on Fri, 02/12/2021 - 12:22
EE, Thanks for the detailed walk-through of the rationale. It has really helped. I shall print this out and keep it inside my copy of EE. Certainly a large part of my consternation is trying to fit EE citation principles into the citation manager tools that the various genealogy programs offer. I would not be taking the approach of a lengthy bibliography for each record used. But where I *am* able to properly identify an online record and therefore cite it in layer one of the reference note, in at least some cases I would plan that my bibliography would reference the database collection it was found in to cover the multiple records from that source. Does that make sense? Thanks, Jeff