Degrees of Freedom Analysis for Genealogists

Often Degrees of Freedom (D of F) is used when solving algebraic math problems, however I have found that it is a particularly useful way of searching for and analyzing records. In math terms D of F has the following generic formula:

Number of Total Variables – Known Variables = D of F

For genealogists the above formula applies when the number of variables is viewed as the types of information that are present in the record, and the number of known variables are the information that we know about the individual. Similar to a math problem, having more known variables than the total of variables results in an over parameterized problem. You will have conflicting information or potentially no results at all. This is one of the most common mistakes made when conducting a search because it is tempting to fill in all the boxes with what we know even if all that information has nothing to do with what we are looking for. On the other extreme is an unspecified search where there are more than 1 D of F. Although unspecified problems aren't unsolvable they are much more challenging as you often have to resort to trial and error. Sometimes with genealogy you get lucky and the less specified search reveals a small handful of records, other times you aren't so lucky and it yields thousands. Remember: a vague question is often met with a vague answer! The best way to avoid doing either of the above is to conduct each search specifically aiming to answer ONE SPECIFIC research question. By understanding what type of record we are searching for and what information it contains we are able to adjust our search parameters accordingly – ultimately aiming for that 1 D of F.  

Now lets take a look at an example of how I applied the above technique when trying to locate my ancestor in the England and Wales Civil Registration Birth Index. The research question I was trying to answer with this search was: which borough of London her birth was registered in (as that was likely where she was born). I have conflicting information about her birth, having seen her claim to be born in Blackfriars, Southwark, and Bow – with Blackfriars being the most consistent during the later years of her life.

Birth indexes supply the following information: child’s name, year of birth, yearly quarter when they were born, county of registration, and borough of registration - hence 5 variables. I know her name was Bessie Sword and that she was born June 5 1875 somewhere in London – hence 4 known variables. (June falls in the 2^nd quarter which covers the months of April, May, and June). When I sub the above information into the D of F formula I find that I have 1 D of F being the borough where her birth was registered.  

5 variables – 4 known variables = 1 D of F

I only entered the known information into Familysearch.

Notice that I DID NOT enter any marriage or death information. I DID NOT enter any relationship or residence information. I ONLY entered the known information from the D of F analysis. As it turns out there was only one record that matched my search for that collection.

After analyzing the record, I concluded that this indeed was the person I was searching for and thus was able to answer my research question.