For some reason I am struggling to understand different explanations about this. I have been told that a Census is an event, and the sheet of paper is a source. This is similar to a Birth being an event, and the Birth Cert being a source. Is that correct?

So do I add a single Event of the, say, 1901 census, then what do I add to each person? Another event?

  • If you put it in as a census and you get residence info that differs in some way for the same time period, it will be easier to notice. – user5344 Jul 22 '16 at 17:49

Part of the reason for confusion here may be that genealogy software tends to be very person-centered, and it's easy to fall into a person-centered workflow. We tend to think of gathering information about a person, then adding it to our database, then attaching the source citation as an afterthought.

It's natural to take a person-centered approach as we are looking for sources -- the reason we collected it in the first place is because we thought it might belong to one of our people. But once we roll up our sleeves and analyze the source itself, it's important to switch things up and learn more about the nature of the source itself, so we can understand what the source is telling us.

Principles of Evidence Analysis

If you look at the Evidence Analysis Process Map from Elizabeth Shown Mills' Evidence Explained (see QuickLesson 17: The Evidence Analysis Process Map), you'll see three critical elements mapped out: SOURCES, INFORMATION & EVIDENCE. Each of these three elements can be broken down into three different classes, which is why you'll see Mills' map referred to sometimes as "the 3 x 3".

In a source-focused approach, you start with a source, create the source citation for the source first, and then extract all the information inside the source, analyzing it as you go. There are two basic reasons to do this -- first, to understand the nature of the source, and second, to remember where the information comes from (both for ourselves, and so we can point other people to the source if they need to check our work or find the source, too).

The source (document, etc.) is the container -- the information is the contents. For evidence, I'll quote Mills rather than paraphrase:

Evidence is our interpretation of information that we consider relevant to a particular research question.

George G. Morgan gives a good example of working from a source in his blog post Three Genealogical Exercises -- see the section Dissecting Obituaries:

First, I read the obituary in full. Next, I use a pencil to underscore individual pieces of information in the obituary that point to some resource that may or will be of genealogical value. The obvious items are name, gender, age, residence, life events, place where a funeral or memorial service is scheduled, names of officiating clergy, place of interment, and names of any survivors. Other information may include occupation, name of spouse(s), sibling(s), place of birth, life events, military service, church affiliation, occupation, and more.

I prepare a list that includes each and every one of these underlined clues. Underneath each one, I notate 1) what information that clue can provide; 2) what records might exist of the fact or clue; and 3) where the record(s) would be held. A typical short obituary usually has at least a dozen such clues to records. I then use telephone and city directories, the Internet, and other resources to determine the location that I would contact for more information. The exercise takes about thirty minutes or less for each obituary.

To do the same analysis for the census, for US Census records, I often use these links as a prompt:

There is much more information that can be gleaned from a census record, especially from the really nosy 20th-century US Federal census returns, than simply recording the basic information of name, age, location, etc. In a source-centered approach to research, whether to record a source as a source is never a question -- sources always get recorded, and if you use the information in them, you always record where the information came from.

That being said, asking if you should record an event for a census is a good question. But -- which event to use?

Census versus Residence

I haven't used Gramps much yet, but to give you an example to think about, I'll use a problem that comes up often when talking about census records -- if you record them as an event, do you record them as a Census event or a Residence? When users attach a census record in Ancestry's online tree system, the tree system creates a GEDCOM residence event. I prefer to create an event for the census for each person in the household, and I choose the census event for the following reasons:

  1. The instructions for what information was recorded in the census varies from decade to decade and from country to country. For example, in the census of England and Wales, the enumerators were asked to record people who were visitors to the household on census night. If this census is not someone's primary residence, putting in a "they were here" date-and-timestamp for the census as a residence is misleading. It is not their residence, and by giving yourself an erroneous impression that they lived there, you could block yourself from further discoveries.
  2. If one of your goals is to locate all available census records for every person in your research, it's much easier to spot that you've found them all if they are actually in your database as a census event. Entering the census as such allows me to make use of Family Historian plugins that help with running routine queries, GEDCOM checkers, or other research assistants such as GenSmarts, which are designed to look for missing census records.

For Family Historian, there is an add-on program called Ancestral Sources which allows the user to do enter all the people in the household at one time. This is much easier than creating one census event for the head of household and then trying to copy-and-paste it, modifying it for each person. AS creates the Census event (and birth events as needed) for all the people in the household, much like the Form Gramplet in Gramps already mentioned in this answer.

After entering the basic information in AS, Family Historian's Auto-Source Citation feature allows me to easily select the census from the source list, and add other information from the census (such as the information listed in the Clues in Census Records articles) which was not contained in the Ancestral Sources template. The Auto-Source Citation automatically flags each piece of information I enter with the source citation as I do the data entry.

To me, setting up the source first, working from the source, and extracting the information, with Family Historian putting in the source citation as I go, is a lot easier than the person-oriented workflow of putting all the information in first, then attaching the source citation afterwards.

Even if you can't replicate my workflow in Gramps exactly, you'll understand the information in each source better if you learn to examine sources critically and enter the source as a source into Gramps first, before you do the data entry for the information inside.

Previous Qs that touch on the topic of evidence management and source-based genealogy:


There are several ways you can approach this. You can consider a census as an event, even though it was not an important life event. Since it was not an important event, you would also not be wrong to only consider a census a source.

In Gramps, one way of doing it is:

  1. Add the census as an event from the Person's main page: enter image description here enter image description here

  2. On the census event, go to the Source Citations tab, and add the necessary details. I am a minimalist on this; I only include the information needed to find the record, not every extraneous detail that you find in a citation generated by Ancestry.co.uk. enter image description here

  3. Under the citation, you can include Notes, such as a transcription. enter image description here

  4. You can add this same event (including all its citations) to the other people named in the record. You can do this by using the "Share an existing event" button that looks like: enter image description here

There are other equally good ways of doing this. You can add all the censuses to the Source Citations on the person's main page. This makes it quicker and easier to see which sources you have added. You can then add them to the relevant events if needed.


In general, there is no wrong way to use Gramps... do whatever makes sense to you. But for census data, there is an Addon (third-party plugin) called Form Gramplet that will add the information in a standard way so you don't have to think about it.

I suggest you try it, and see if you like the way that it organizes the data. You can update your addons from the Preferences: menu -> Preferences -> General tab -> Check now


Hers's how I organize what I see on a a census page:

  • Source: the 1900 US Federal Census.
  • Citation: lines 40-44 of Page 6B of Enumeration District 44 of Atlanta, Fulton County GA.
  • Census Event: the people referred to in the citation at a specific point in their lives.
  • Multiple Persons share that Census Event.
  • Birthday Events use the Citation as evidence.
  • Occupation Events also use the Citation as evidence.
  • "Use the Citation as evidence" makes no sense to me. I realize that this is an artifact of how our software (not just Gramps) is using the term "Citation". Sources contain information, which becomes evidence when we use the information to answer a specific research question -- see Evidence Analysis Process Map. All of your events cite the source as the container for the information. To me, a Citation is the statement of where you found the information. It's a pointer to information, not the source. – Jan Murphy Dec 1 '20 at 17:45
  • @JanMurphy there might be terminology differences between Gramps and EAPM. Personally, the Gramps usage seems reasonable to me, because it fits with how you use the term citation in a school research paper: it's a paragraph from a "book" (or line from a census page) which is evidence for your argument, while the source is the "book" that the citations come from. Thus, a source can produce multiple citations, and each citation can be used as evidence for multiple events. – RonJohn Dec 1 '20 at 17:59
  • To clarifiy: The paragraph itself from the book which you are citing is an extract from the book. The citation is the footnote saying where the paragraph comes from in the book. The citation, which is what would appear as part of the bibliography for your school research paper, doesn't have any information from the source in it. Another researcher can't use your citation as evidence; they would have to find the original source using that information & read/analyze the source itself. The principle is to cite what you use. – Jan Murphy Dec 2 '20 at 1:16
  • @JanMurphy "The principle is to cite what you use." How am I NOT doing that? The research paper footnote says what page the text is from, and points to the bibliography, while the citation "lines 40-44 of Page 6B of Enumeration District 44 of Atlanta, Fulton County GA" from Source "1900 US Federal Census" leads anyone with two functioning neurons back to the original. – RonJohn Dec 2 '20 at 4:08
  • 1
    @RonJohn - the issue is not what you are doing - the issue is the names that you are using. It's like the difference between an address and a house. The address describes the house, but it's just a string of text - you can't actually live in a string of text. Similarly here - the actual evidence is in the source (the text is in the book). The citation only points to where you found the source and how you recognise it. (The citation contains the book's title, publisher and perhaps the library). The actual text is the evidence and that's in the source. – AdrianB38 Dec 4 '20 at 20:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.