A genealogist is a special kind of detective.
Just like a police detective you must unravel mysteries and sort through conflicting bits of information. A police detective has the advantage of interviewing eyewitnesses whereas the genealogist must deduce everything from old documents and hearsay. Both must be able to distinguish between fact and fiction.
Part of being a good detective is looking at your evidence and deciding what is feasible and what is not. This is an important and necessary skill if you want to be a good genealogist.
If you fount this family group on the 1880 census, what would you deduce?
David Smith, age 50
Mary Smith, age 25, wife
John Smith, age 15, son
Elizabeth Smith, age 13, daughter
William Smith, age 8, son
Michael Smith, age 6, son
Sarah Smith, age 1, daughter
Did you notice the 15-year difference in the age of David and his wife Mary? Son John and daughter Elizabeth are too old to be Mary's children. David was most likely married previously and that would be an avenue to pursue. However, you must also consider that even though John and Elizabeth are listed as son and daughter, they could be nephew and niece. Maybe David is raising his dead brother's children. The other three children appear to be David and Mary's. Again, that might not necessarily be the case.
David is old enough to have grandchildren. If he has a 15 and a 13-year-old at home then he just might have older children that are no longer living in the house. Is Sarah David's daughter or is Sarah the daughter of a dead older son? If you believe the youngest three to be David and Mary's, why is there a 5 year gap in age between Michael and Sarah? Could Mary have lost one or two babies? We could even go further and say that maybe none of these children are Mary's. Maybe she and David had just married and all of these children belong to his first wife.
You need to formulate a research plan to check out each of these possibilities and eliminate them one by one until you are left with only one possibility. Your biggest mistake would be to write this information in your family file at face value without doing any further research.
Looking at one census year is not enough. You need to follow this family back in time. Try and find David's family in the 1870 census first. Then go to the 1860 and the 1850. You always want to work backward in time, not forward. When did David actually marry Mary? That one piece of information will shed light on several of the dilemmas. One of your first moves will be trying to locate David and Mary's marriage license.
You can not take anything at face value when doing genealogy research. You must investigate every possibility. Like a police detective you must look at every piece of evidence from all angles before coming to any conclusions.