Category: Census

Post 17 left us with Lafayette Braddock’s Birth plan framed and saved. Subject picked, claim type set, a research question on record. Now we click “Continue to Gather Evidence,” and it opens the research log itself. We don’t start from a blank page here. Evidentia often puts one row in the grid before we add anything at all.

What’s Already Waiting for Us

When we open Gather Evidence, we’re not always starting from scratch. One row here was already marked Evidence. Its “Claims that…” column held text we’d recognize right away: “the father of Hayward, Essie M was born in Canada (Eng).” That’s the same claim the Form Assistant produced back in Post 16, showing up here on its own. When Evidentia already has evidence linking a subject to a claim, Gather Evidence pulls that row in for us automatically. We don’t have to re-enter or re-transcribe anything. Classification still reads Indeterminable here, same as it did in Catalog Claims. One accepted claim isn’t proof by itself. It’s a fact worth weighing.

Notice that this row has no Repository listed. It came from a transcription, not a fresh search, so the grid doesn’t invent a repository it was never given. That’s worth remembering as we work: an auto-filled row only knows what we’ve already told Evidentia elsewhere.

Adding the Searches We Still Owe

One row of evidence answers “was he born in Canada.” Post 16 already established that. It says nothing about where in Canada, and that’s the actual question this plan exists to answer. This is exactly where we want to add rows of our own. Genealogists have kept research logs for this exact reason, long before software did it for them. FamilySearch’s own guide to research logs makes the same point: we should log every search, whether or not it turns anything up. So we click Add Row twice, once for each search we still need to run. One row becomes an 1881 Census of Canada search for Lafayette Braddock. The other becomes a 1900 US Census search in Carver, MA, a different town than the Halifax household Essie’s record came from. That second search is worth running on its own.

Both new rows show Status: Planned. Nothing’s been found yet, so there’s no Evidence badge here, just a logged lead. We set Repository to ancestry.com for both. The “Claims that…” column still holds its italicized placeholder, “Record findings relative to research question…”, and Classification sits on “Select…” None of that is filled in yet, because nothing has been entered yet. A research log that only tracks what we’ve found misses half the point. The leads we haven’t run down yet are the whole reason a research question stays open.

A row that comes back empty just stays Planned, with nothing to classify yet. A row with contradictory findings gets the same Classification dropdown as everything else, plus a Notes icon to explain the conflict.

One Grid, Several Jobs

The toolbar above the grid does more than add rows, and the icons on each row do more than we might notice on a first pass. Here’s what we can actually do from this one screen:

• Add Row. Log a new search before we’ve even run it, so it’s tracked the moment we think of it.
• Create Worksheet. Build a printable version of this same log, handy for an archive visit where a laptop isn’t practical. That’s its own post later in this series.
• Research Notes. Keep notes at the plan level, separate from the small notes icon on each row.
• Columns. Hide whichever fields we’re not using for this particular plan.
• Document a nil result. When a search comes back empty, we can write that directly into the “Claims that…” field, something like “no record found in the 1881 Census of Canada,” instead of leaving the row blank indefinitely.
• Save to Claims. The bookmark-style icon on a row turns its findings into a real Claim once there’s something worth saving.
• Add a note. The notes icon on a row holds a longer explanation, including how a conflict got resolved.
• Delete a row. The trash icon removes a row we no longer need. It’s grayed out on the Evidence row, since we don’t delete something already backed by real evidence.

We even have a legend for these icons, tucked under “What do these icons mean?” above the grid. Save-to-claims is enough on its own to fill the next post in this series, so we’ll save the deep dive for then.

Three rows sit in Lafayette Braddock’s Birth plan now. One fact is already on record. We still need to do two record searches. We don’t have to keep it in our heads or dig through old notes to remember this.

Next in this series, we’ll look at what happens when one of those Planned rows actually turns up something. That’s Save to Claims, and it’s how a row in this grid becomes a real Claim record.

— Ed Thompson, Evidentia Software

Post 16 ended with a research question I hadn’t actually written down anywhere: where in Canada, specifically, did Lafayette and Mary Braddock come from? A second one came with it — who are the other four children the 1900 census says Essie had, when I’ve only got one of them on file? Writing those down in my head was easy. Doing something with them is a different problem — and it’s exactly the one the new Research Plan is built to solve.

Evidentia 4 ships Research Plan as a genuine three-step workflow: Frame Question, Gather Evidence, Analyze Evidence. This post is Step 1 — the part where a vague “I should look into that” turns into an actual, named research question attached to an actual person. I picked Lafayette Braddock, Essie’s father, since his Canadian birthplace is the loose end Post 16 left hanging.

One complication: Lafayette isn’t in the database yet. He’s been a name on Essie’s family group sheet for years, nothing more. Starting a research plan for someone who doesn’t have a Subject record turned out to be its own small lesson.

Opening Research Plan

Research Plan lives in the main toolbar, not tucked under a menu. It’s the small notepad-and-checklist icon next to the star — easy to walk past if you’re not looking for it, since nothing about it says “new feature.”

Clicking it opens a fresh plan, ready for a subject.

A Subject Who Isn’t Here Yet

Lafayette isn’t an existing Subject in my database — just a name on Essie’s family group sheet, never entered as a person in his own right. Rather than backing out of Research Plan to go build him first, there’s a person-plus button right beside the Main Subject field. Clicking it opens an Edit Subject modal without leaving the plan.

It asks for the name twice, in two different shapes: “Braddock, Lafayette” for how he’ll sort in lists, “Lafayette Braddock” for how he’ll read in a report. Gender, a Ref No., and a FamilySearch ID are all optional. I filled in the two names and gender, left Ref No. and FamilySearch ID blank, and clicked Done. Lafayette Braddock existed as a Subject for the first time, without a single screen change.

Framing the Research Question

With a subject picked, Frame Question fills in the rest of itself. The Plan Title became “Braddock, Lafayette: Birth” automatically — subject name plus claim type, no typing required.

Claim Type is where I expected to pick “Birth Place” and reach for the dropdown to find it. It isn’t there. Evidentia’s claim types just have “Birth,” and it covers date and place together — which is actually the more honest way to frame it, since a birth record rarely gives you one without the other. I chose Birth, and the Research Question field filled itself in to match: “When and where was Lafayette Braddock born?”

The auto-suggested question is a starting point, not a final answer — the field is editable, and the census already narrows things past “born somewhere.” Lafayette wasn’t born just anywhere; the 1900 form says Canada. The plan captures the general question now; the specific one — where in Canada — is what Gather Evidence and Analyze Evidence are for.

Writing the question down first isn’t just an Evidentia habit. The Board for Certification of Genealogists’ Genealogy Standards call for a clearly stated research objective before the record-pulling starts — a research plan just gives that objective a place to live.

One Subject, or Two

Birth only ever needs one subject — a person either has a birthplace or doesn’t, and there’s no second party to a fact like that. Not every claim type works that way. Marriage, and parent-child claim types like Parent, are inherently about two people, and Frame Question adapts: pick one of those and a second-subject field appears next to Claim Type.

I tried it two ways, on the same person.

First, Marriage. The second-subject field is labeled “Spouse / Partner,” and its placeholder — “— Select person (optional) —” — means exactly that: optional. I left it blank, and the auto-suggested Research Question fell back to something generic: “Research Marriage for Lafayette Braddock.” Not a real question yet, just a placeholder worth typing over — which is fine. A relationship claim type doesn’t force you to already know who the other person is before you can start a plan.

Then I picked Parent instead, and added Essie Braddock as the second subject. The field label changed to “Child” — Evidentia already knows which side of a parent-child relationship I’m asking about, based on the claim type and direction, and labels the field accordingly instead of making me guess. A checkbox appeared underneath: “Limit evidence to this specific pair only,” checked by default. With a real second subject in the picture, the Research Question sharpened too: “Was Lafayette Braddock the father or mother of Essie Braddock?” — an actual yes/no question, not a placeholder.

That checkbox is the real flexibility. Checked, the plan only looks at evidence linking Lafayette and Essie specifically — useful when you already know the pair and want to confirm or disprove it. Unchecked, it would widen to any evidence of Lafayette in a parent role at all, useful when you’re not sure which of his several children a given record is even about, or when you’re researching him as a parent generally rather than one specific relationship. Same claim type, same main subject, two different research questions depending on how much you already know.

No Stepper, Just Three Tabs

I’d expected a progress bar across the top — numbered circles, checkmarks filling in as each step gets done, the kind of thing plenty of wizards use. What’s actually there is plainer: three text tabs, Frame Question, Gather Evidence, Analyze Evidence, with the active one underlined. No numbers, no checkmarks, no lock icons — just plain labels. It’s a smaller piece of UI than I was picturing, and it does the job without getting in the way.

Clicking “Continue to Gather Evidence” moves the plan forward — Lafayette Braddock’s birth question, framed and ready — into Step 2.

That’s where this series picks up next: actually gathering the evidence to answer where in Canada Lafayette Braddock was born.

Research Plan is in the main toolbar now, next to the star, whenever a loose end is worth framing properly instead of just remembering to get back to it. Next in this series: Gather Evidence, and what it actually takes to answer where in Canada Lafayette Braddock was born.

— Ed Thompson, Evidentia Software

I didn’t know where Essie Braddock Hayward’s parents were born until I transcribed one row of a census form.

This is a full Form Assistant walkthrough — one real record, screenshot by screenshot.

Lafayette Braddock and Mary E. MacHenry Braddock have been in the database for years, attached to their daughter Essie as names and nothing else. Then I opened the 1900 census for the Hayward household in Halifax, Massachusetts, ran it through the Form Assistant, and two of the suggested claims were about people who aren’t even listed on that page: Essie’s father and her mother, both recorded as born in Canada.

This is the walkthrough Introducing the Form Assistant promised — from the citation to the claims sitting in Catalog Claims. No placeholder household. This is Essie’s actual family, transcribed off her actual 1900 census page.

Opening the Citation

I already had the citation sitting in Catalog Claims from earlier research: 1900 US Census – Halifax, MA – Edwin S Hayward. I opened it and clicked Form Assistant, next to Mini-Editor.

From there it’s a list of every census form Evidentia knows how to read, 1790 through 1950. I picked 1900.

Filling In the Form

The header fields come first, copied straight off the image: Massachusetts, Plymouth County, Halifax. Supervisor’s District 115, Enumeration District 1123, enumerated 11 June 1900.

Then the household itself — three rows. Edwin S. Hayward, head, 47, a farmer, born March 1853. Essie M. Hayward, his wife, 30, born August 1869. Their daughter Mabel G., 11, born March 1889. Straightforward transcription: name, relation, sex, birth month and year, age, marital status, years married, children born and living.

What Suggest Turned Up

I clicked Suggest expecting the obvious — her birth month and year, how long she’d been married. Both came back, and I accepted them without much thought.

What stopped me were two claims I hadn’t asked for: the father of Hayward, Essie M was born in Canada (Eng), and the mother of Hayward, Essie M was born in Canada (Eng). I’ve had Lafayette Braddock and Mary MacHenry Braddock in the database for years — parents on a family tree, nothing more. I didn’t have where either of them was born. The census did, sitting in two columns past the ones I was actually paying attention to, and the Form Assistant turned them into their own claims instead of letting me transcribe past them.

(“Eng” isn’t a typo. Canada wasn’t an independent country in 1900 — it was a self-governing Dominion under the British Crown, and full legislative independence didn’t arrive until 1931. So the census didn’t ask for a nationality; the 1900 enumerator instructions told census-takers to record English or French Canadian descent instead, which is exactly what landed in this claim.)

Accepting the Claims

Accepting is one click per suggestion, and I took every one that came back for Essie’s row.

Back in Catalog Claims, there they were: five new Claim records for Essie — her birthplace, her years married, her father’s birthplace, her mother’s birthplace, and her birth month and year — sitting alongside the ones already recorded for Edwin. Every one of them is currently classified Indeterminable, with nothing filled in under “provides evidence for.” That’s exactly right. Accepting a suggested claim doesn’t prove anything by itself. It gets a fact out of the image and into a record you can actually work with. Deciding what each claim is evidence for, weighing conflicts, reaching a conclusion — that’s Analyze Evidence’s job, the way the Genealogical Proof Standard describes it, not the Form Assistant’s.

So what did one row of a hundred-and-twenty-six-year-old census form actually settle? Essie’s birthplace, already in the file from a Find a Grave memorial, is now backed by a federal record instead of a headstone transcription. Her parents’ birthplaces weren’t documented at all, and now they are.

It also left me with two questions I didn’t have before I sat down: where in Canada, specifically, did Lafayette and Mary come from — and who are the other four children this same row says Essie had, when I’ve only got one of them, Mabel, on file? Those are exactly the kind of open threads the new Research Plan is built to carry forward, and that’s where this series picks up next.

The Form Assistant is waiting in Catalog Claims whenever you open your next citation. Next in this series: framing an actual research question — starting with where Lafayette and Mary Braddock were really born — in the new Research Plan.

— Ed Thompson, Evidentia Software

I built the Form Assistant because a census page hands you a dozen or more facts about one household — names, ages, birthplaces, occupations, relationships, sometimes military service or literacy — and cataloging every one of them by hand, one claim at a time, for every person on the page, is real work. I’ve caught myself skipping facts when “data fatigue” sets in, the ones that feel too minor to type out on their own. Researchers have been doing it that way since Evidentia’s first release.

Open a citation in Catalog Claims, click Form Assistant, and you get a table that matches the record in front of you, one column for each fact the form actually asks about.

All 18 Census Years, One Assistant

Every U.S. federal census from 1790 through 1950 is supported, including the 1860 Slave Schedule. Pick the year printed on your record and the Form Assistant loads the columns for that year. An 1850 household schedule doesn’t ask the same questions a 1940 one does, and the form in front of you shouldn’t pretend otherwise.

Transcribe, Then Let Evidentia Suggest

Fill in the rows the way you’d read the page: left to right, person by person. Each row saves as you go, so a closed laptop or an interrupted library visit doesn’t cost you the transcription. When you’re done, click Suggest. Evidentia reads back everything you entered and proposes a claim for each fact you captured, ready for your review. Arrow keys move you up and down a column the way they would in a spreadsheet, so your hands don’t have to leave the keyboard between rows.

Review Before You Accept

I don’t trust a suggested claim blindly, and you shouldn’t either. The Form Assistant only knows what you typed, and what you typed can still be wrong — a census taker’s sloppy 3 read as an 8, a row filled in from the line above by mistake. So nothing gets added to your citation without you looking at it first. Accept a suggestion and it becomes a real Claim record. Reject one and it’s off the list, at least until you change your mind — Unreject brings it right back.

Edit a row and click Suggest again, and I didn’t want your earlier calls thrown out every time the page gets re-checked. They aren’t. Whatever you already accepted or rejected stays exactly as you left it.

From there it’s the Catalog Claims you already know: pull the claim into Analyze Evidence, weigh it against the citation, and write your conclusion the way the Genealogical Proof Standard asks you to. The Form Assistant doesn’t skip that step for you. It was never supposed to.

The next post in this series walks through a real census page from start to finish, one household, every suggestion reviewed. Until then, the Form Assistant is waiting in Catalog Claims whenever you open your next citation.

— Ed Thompson, Evidentia Software