Creating a "Census Candidates" Search List I

"I want a list of people in my database for whom I have no 1950 census record." That seems like a pretty simple project, right?

• The name itself tells us what TMG report we probably want: a List of People report.
• We don't want a list of all the people in the database. We want a filtered list of people.What people belong in this report? The subjects of this report must meet two criteria.

1. They must have been living on 1 April 1950, the official census date.
2. The database must be missing 1950 census information for those people.

I last wrote about this report in 2005, "The 'Census Candidates' Report." Twelve years is an eon in computer years, but the process and principles in this article still apply. What I discovered, though, is that my research work flow has changed in this time period, and I seldom create this census candidates report as presented.

1. I virtually never print out a list of census candidates and take it to the library or archives. With so many census records online, most of my census research is done in my bunny slippers - and it's done incidental to a problem on which I'm working.
2. I do create a census candidates report to test my progress on a given problem, though. That's the report I will discuss here.

Take a look at the filter in line one. To determine whether or not a person was living on the official census date, a researcher must know something about that person.

• This is easy, if a person's birth and death dates are both known.
• If the birth date is known, but the death date isn't, applying a probable life span would suggest a person was living at the time the census was taken.
• With neither birth nor death date known, one could determine probable candidates by filtering for those with any events falling within a given span.

The Arlington RUG has published several articles on TMG filters. At least two of them include discussions of census candidates filters: "Favorite Filters" and “1880 Census Candidates Using Flags and Book Manager.” For those of you who have created unique census tags for each census year or who want to identify meticulously all people in your database who might be living at the time a given census was taken, these discussions should be read and studied. I don't fall into either category. (1) If I created unique census tags for each census year in my database, I would currently require 52 of them. (2) If I don't have at least a say date for a person in my database, I'm not interested in that person.

My census candidates research focuses on those people for whom I have both birth and death groups tags. Note that either date may be estimated, but there is a tag for the individual. Census research is family research, not individual research. I've discovered that when I find one known person in a census, his or her family usually includes individuals for whom less information has been found. To illustrate this statement, I created an 1850 census candidates list for a current project. That list included 25 names for whom both birth and death dates are known or estimated and 27 names for whom only birth group dates are known. I worked through the list of 25 names, and when that research was complete, the 27 names on the second list had shrunk to only eight.

My census candidates report uses (1) Book Manager, (2) three sequential reports that set my Temporary flag values, (3) the creation of a Focus Group that includes only those individuals included in a current research project, and (4) the Project Explorer filtered for those Focus Group individuals with a designated Temporary flag value. The next post details the process.