Monday, September 16, 2013

The research three-step

Back in the day, my basic research plan was the same for every problem: look for whatever on the internet, download it, and dissect its parts into my genealogy database. This can work -- not all genealogy questions are difficult to answer.

But often it doesn't work. Or it gets a wrong answer. In that case we need to step back and make an actual plan. I suggest three steps:

(1) Ask a particular question about a particular known person in a specific place and time. Not "I want to know all about Grandma." But something like, "Who were the parents of Tryphena Burton?" or "Did she have a sister Tryphosa?"

(2) Look for the records that may tell you the answer in so many words ("direct evidence"), such as a birth or death certificate, a marriage application, a pension application, or a compiled genealogy. The exact records you can hope to find will vary a lot from place to place and time to time. One checklist I like just runs through a basic dozen common categories: church, cemetery, census, compiled, court, land, military, naturalization, newspaper, probate, tax, and vital. There are plenty of others, and this is where it pays to become well acquainted with each locality.

Often step 2 produces records that disagree with each other. That's another post one of these days.

(3) If these and other sources don't tell you the answer and don't even supply you with a nice juicy contradiction, then start looking for records that will give you clues. Actually you can start by looking again, more closely, at the records you've already collected. This is when some record types, like court and probate and land records, come into their own. They can show proximity and association and help you begin to sort out all the Burtons in the area, and identifying who were friends with which branches in which parts of the county.

In real life these steps may overlap. More experienced researchers will take the time to see more in any given record, and get more than the directly stated evidence from the start. But they won't assume that the answer requires prodigious effort in accumulating clues either, and skip right to that step.

Harold Henderson, "The research three-step," Midwestern Microhistory: A Genealogy Blog, posted 16 September 2013 ( : viewed [date]). [Please feel free to link to the specific post if you prefer.]


Elizabeth Lapointe said...

Hi, Harold,

Yes, I agree with the research three-step, but don't forget to give each 'find' a certain weight.

For example, I would conclude that the age given in a census report and an age calculated by examining a birth record are of different weights. Am I right about this?


Harold Henderson said...

Elizabeth --

I agree in general, but every case is different. (One could imagine a string of census reports all agreeing and being in serious contradiction with a birth record. Then we get into correlation as well as analysis.) The best article I have read to make us remember such possibilities, and not be too dogmatic about "better" sources, is Tom Jones's NGSQ article of some years back, "The Children of Calvin Snell." Don't miss it!