Disproving a Hypothesis and Moving the Research Forward
What do you do when your research hypothesis is disproved? Perhaps the discovery of a new record, lack of DNA evidence, or even discovering another person of a similar name could lead to your proposed ancestor turning out to be incorrect. Do you give up or use the findings as an opportunity to move forward with a new avenue of research? The silver lining in this scenario can emerge with a reexamination of the research question, record analysis, and a new research plan.
I recently let go of the hypothesis that George W. Dillard was the father of my ancestor, Cynthia (Dillard) Royston. If you have followed my research, you know how much work has gone into this effort. Two Research like a Pro study groups in 2018 brought together my research of the past 15 years about George and identified and eliminated other Dillard men as candidates for Cynthia’s father. Even after completing two projects doubts lingered and I could not state with confidence that George was the father of Cynthia. I had not connected George to Cynthia on my Ancestry tree and Thrulines showed another possible family line for her adding to my doubts.
Disproving a Hypothesis
Recently, a podcast listener and proven descendant of George W. Dillard contacted me via Family Locket. She had listened to RLP podcast episode 10 , the George W. Dillard Case Study, where Nicole and I discuss the Dillard research. Continuing our correspondence through email, this listener shared images of the Dillard family bible where the female born in 1815 was identified as Mariah L. Dillard married to a James [K—-]. Mariah’s headstone further named her as M Louisa Dillard, wife of James Kivlin. With only one female in the 1820 and 1830 household of George W. Dillard of an appropriate age to be Cynthia, my hypothesis was disproved. Mariah Louisa (Dillard) Kivlin was the daughter of George W. Dillard, not my ancestor, Cynthia (Dillard) Royston.
To further test this new information via DNA, I added George W. Dillard as the father of Cynthia to my Ancestry tree then waited for Thrulines to populate. After a few days, only a few matches appeared, those that were known descendants of Cynthia who had added George as the father based on the circumstantial evidence. With no DNA matches to the many known descendants of George W. Dillard, I discarded my hypothesis.
Reexamining the Research
My research question remained, “who was the father of Cynthia (Dillard) Royston born about 1815 in Georgia and died 2 August 1882 in Collin County, Texas.” A previous research project had identified ten candidates who fit the requirements for a father/mother of Cynthia:
-Born before 1795
-Residing in Georgia in 1820 with a female about five years of age in the household
-Residing in Georgia, possibly near the Alabama border in 1830 with a female in the household about fifteen years old
The research attempted to disprove each of the ten candidates via probate, marriage, and census records as well as family trees. All nine men were eliminated leaving only one woman, a Susan Dillard of Muscogee County, Georgia, as a possible connection. Her 1830 household revealed a female of appropriate age for Cynthia and although it is unlikely Susan was Cynthia’s mother based on the age difference of only fifteen years, it is possible.
Reviewing the research led to several new questions. Could Susan be the second wife of Cynthia’s father who was likely deceased by the 1830 census? George W. Dillard was also residing in Muscogee County, Georgia. Was he an extended family member, perhaps a cousin? Could Cynthia have been in the 1830 household of a Dillard man residing in Alabama instead of Georgia?
Moving the Research Forward
With George W. Dillard eliminated as the father for Cynthia, I can now advance my research via DNA as well as pursuing a new hypothesis. Having my research logged and written made it easy for me to review my findings and explore new avenues for the future without having to wade through unorganized research notes – a huge benefit of following the research process. If you’d like to know more about this case, you can read my reports of the two research projects centered on this research question.
We invest a great deal of time in our research and we may be guilty of what is known as “confirmation bias” – trying to make the pieces fit together based on circumstantial evidence or what we want an answer to be. Realizing that a key piece of evidence may emerge, such as the Dillard family bible, we do our best to form and test our hypotheses. When our theory is disproved, rather than feel disappointed, we can be excited to finally put that theory to rest and go to work on a new avenue of research.
Best of luck in all your genealogical endeavors!