In my lectures, I often joke that I can identify an Irish family researcher by the two bald spots on the sides of his or her head--a result of pulling out hair in frustration over trying to resolve age discrepancies in the census and other records! For a long time, I thought that my ancestors' age variations were simply a Large family trait. My great-grandfather had married a much younger woman, and so I concluded that his "finding the fountain of youth" and aging backwards between 1860 and 1880 was an effort to appear to be closer in age to his wife. The other explanation could be that one of the children gave the information to the census taker, and simply presumed that ma and da were about the same age.
But most Irish family researchers I meet describe the same problem: a variation of from five to even fifteen years in age as reported in the US census.
John Grenham, in his book Tracing Your Irish Ancestors, Third Edition (Genealogical Publishing Company: 2006) (a "must-have" volume for your personal genealogical library) recommends "a large dose of scepticism is necessary" when working with dates of births, marriage, and deaths before 1900 (page 7). Grenham notes widespread discrepencies even as late as between the 1901 and 1911 Ireland census. He attributes the discrepancy not to vanity, as I had assumed with my great-grandfather, but to people in the 1800's not knowing their date of birth. I myself believe another explanation--this is a joke perpetuated by our ancestors so that they can now laugh at us from above as they watch us run in circles trying to pin down their vital statistics.
These age discrepancies can wreak havoc with your research. Many beginners in Irish research, thinking that dates must match precisely for a record to be an ancestor's, pass over non-matching records, or at the least, believe they have hit a "brick wall." At best, encountering an age discrepancy, especially one of fifteen years, adds hours and perhaps years to one's research. After all, there is always the possibilty that the discrepancy denotes a second marriage, a child who died (with a second child given the same name), two siblings (with the same name), or another person altogether. So, I cannot advise you to ignore the different ages reported. Always be aware of these alternative possibilities when researching, and always broaden your research to include years (perhaps even a decade or longer!) before and after a reported birth, death, or marriage.
In my classes, I formerly advised students to use the baptism records as true year indicators (and perhaps even true months, since the Irish baptized very soon after birth). But a glitch in my own research put a stop to that advice. I found my Ellen Large's baptism in November of 1836 in the Roman Catholic church registers in the parish of Castleomer, Co. Kilkenny. Alerted by an old family tale of religious differences in the family, I went to the Representative Church Body Library (Church of Ireland) in Dublin last year. I found the Large children were baptized first at the Church of Ireland, then later in the RC church (I imagine Bridget Kavanagh Large sneaking off to the RC priest with the babies when husband Richard went off to work in the mines). Ellen was first baptized in October of 1833--three years earlier than what I had recorded as her birth year. So there goes that advice out the window!
Perhaps we should lobby the makers of genealogy programs to include special multiple data boxes for the dates of our Irish ancestors' life events. I can see it now, all of the following boxes to replace one single box for birth date: "baptism date" "alternate baptism date" "birth date from 1870 census" "birthdate from 1880 census" "birth date from death certificate" "birthdate from the family Bible" and so on.
Isn't Irish family research fun? Never a boring moment. Love it!