Analysis and Correlation
The third element of the Genealogical Proof Standard (GPS) is “tests – through processes of analysis and correlation – of all sources, information items, and evidence contributing to an answer to a genealogical question or problem.” – Board For Certification of Genealogists, Genealogy Standards, Second Edition (Nashville, Tennessee: Ancestry, 2019), 2.
How does this apply to our genealogical research? After we glean information items and evidence from sources to answer our research objective (discussed in my previous blog post), we apply the third element of the GPS.
Analysis is the physical and contextual examination of a source – a birth register, marriage license, death certificate, family bible, or deed, etc. Why was it created? Who created it? Are any pages torn? Is the item legible? How long after the event was the source created? Could there have been a reason for bias on the part of the creator? Were any corrections made to information at a later date? Were safeguards in place to prevent error or bias?
Correlation is the process of comparing evidence items gleaned from source material. Do all evidence items agree with one another, or are there discrepancies between evidence items?
Elizabeth Shown Mills provides the following categories for sources, information items, and evidence –
Sources are divided into one of three categories –
Original - the deed given to a land purchaser is an example of an original source.
Derivative - the transcript of the bible record I speak of below is an example of a derivative record.
Authored - this blog post is an example of an authored work.
Information items gleaned from a source are placed into one of three categories –
Primary - information provided from someone with first-hand knowledge of an event such as a party to a marriage.
Secondary - information provided by someone with second-hand knowledge of an event such as a husband who was the informant for his wife’s death certificate. The husband would likely not have been present for his wife’s birth, but would have knowledge of when she celebrated her birthday.
Undetermined / indeterminable - information provided by a person not identified in a source. An example – informants for U.S. population schedules are not identified.
Evidence is placed into one of three categories –
Direct - evidence that directly answers a research objective. If the objective is to identify a person’s parents and they are named on this person’s birth certificate, this is direct evidence. This is not to say that direct evidence always provides the correct answer.
Indirect - evidence that does not provide an explicit answer to a research objective, but combined with additional evidence, does.
Negative - when the absence of information is used as evidence.
Below is an example from a personal research project – Amelia Jane (Sellers) Chancellor of Pike County, Alabama: Daughter of Joseph Sellers, Samuel Sellers, or Matthew Sellers –
Through a process of elimination, I narrowed the choice down to three possible candidates for the identity of Amelia (Sellers) Chancellor’s father – Joseph Sellers, Samuel Sellers or Matthew Sellers. I located the clerk’s copy of Samuel Sellers’ Last Will and Testament, the court’s record of his estate’s settlement, and a transcript of the Sellers’ family bible.
Beginning with an analysis of Samuel’s will, I determined that although it is not the original but the clerk’s copy, it is likely free of error. Due to the strict guidelines of the law and quality standards, court records are considered highly reliable for genealogical research.
In his Last Will and Testament, Samuel would have provided primary information – first-hand knowledge – for the identity of his children. Amelia was still alive at the time of Samuel’s death and he most likely would have identified her in his will if she were his daughter. He did not identify Amelia as his daughter. In addition, Amelia is not identified in the court’s record of the estate settlement.
The Sellers’ family bible transcript is an image of a transcript of the original, which does call its reliability into question. A representative of the Daughters of the American Revolution witnessed the creation of the original transcript from the family bible. Given this, the transcript is likely an exact duplicate of the original. Who created the image of the transcript? Was the information in the family bible entered as an event occurred, or were events entered in one sitting at a later date? Without viewing the original bible, this cannot be determined.
When analyzing sources, these are the questions you need to ask yourself.
As part of Amelia’s proof argument, I created the table below – correlating Samuel’s named children according to each source. Amelia is not named in the sources, indicating she was likely not Samuel’s daughter – negative evidence used to help form a conclusion.
Correlation of evidence can take on many forms – tables, maps, narratives, bullet points, etc.
Per the GPS, testing through analysis and correlation is a key step when reaching a sound genealogical conclusion. Are you regularly implementing the GPS when conducting your research?