When doing our research, we must follow the real facts and lead to logical conclusions, rather than just helter-skelter choose the conclusions we happen to like. And it is not a mistaken perception like, just find three records to come up with your fact. You need to find all the necessary and reliable records you need to verify the data in question.
As an example, I have discussed at length before, the birthdate of my great-grandmother was "known" within the family due to multiple records of family listings giving her specified birthdate. But it was until just about 10 years ago when I realized that I had not seen her birth record, that I was apparently the first family member to look that up in nearly 200 years and discovered that her true birthdate was an entirely different date in December, a year later than widely believed and variously recorded on many historical records.
There is information in the standards that you must always be open to finding information that corrects certain facts in your research. Yes, as my example above shows, clearly, I understand that but there are some immutable facts that have been definitively established that anything to the contrary has to be in error. For example, my own birthdate and place is forever set and established because of the evidence that I have. I have my birth record, my baptismal certificate, and the testimony of my late parents and my four living older sisters so therefore the date on my driver license is permanent, set and absolute.
For my dad, his original birth record in rural Upper Michigan was lost in a county courthouse fire. So when he was drafted into the Navy during World War II, his mother and oldest sister attested to his birthdate and that was accepted into his official military service. So therefore the date printed two days off in the Ironwood newspaper is somehow a misprint or misunderstanding because it is contradicted by two firsthand witnesses, these two witnesses being deemed as most reliable.
In doing your own research, begin with your information and then proceed backward in time logically one generation at a time, without skipping "known" generations, to create a reliable, unbroken chain to your ancestry. It has served me very well in my 48 years of my own research into my 10-plus generations of my Finnish ancestry with a smattering of Swedish, German, and Russian forefathers and foremothers.
For all your research, find the document recorded closest in time to the event in question. Birth records, marriage records, and death records, but death records are not as reliable for birth dates because they are the farthest away from the event — birth — and attested to by a potentially unreliable informant, at least someone who was not there at the birth (in nearly all cases).
When convincing records corroborate each other regarding the question at hand, that raises your confidence, but also make sure to explain contradicting information. This will increase confidence in your research by those following in your footsteps when they encounter the alternate and misleading information.
Further information can be found here: http://genealogy.mn/standards-and-ethics-in-genealogy/
*Full disclosure: the gene.mn short-link*, if in this and previous and future articles, is a service of Genealogy.MN, wholly co-owned by this author and my son Abram Isola.