Pre-1880 US census records can allow genealogists to gather clues about family relationships. But at their very essence they confirm that individuals are living in the same household at the point in time when the census enumeration was made.
Ancestry.com’s 1870 census database now infers family relationships from the census enumeration. These two screen shots show that information for Barbara Hays (actually Haase) who is enumerated in Hancock County, Illinois, in 1870.
The inferred spouse is correct and the inferred children are correct are correct in this case. That’s not always the situation for a variety of reasons. The problem in this specific family is that Ancestry.com infers that the children are all Conrad’s children as well–they are not.
Louisa is Barbara’s daughter by a previous marriage and she was enumerated with the last name of her step-father in 1870.
Ancestry.com also indicates that there are seven members of the Hays household.
That’s not quite right either.
There is a thirty-two year old native of Saxony, Charley Walder, living in the household as well. At this point, his relationship to the family is not known.
But he is living in the household and the “household members” list in the “record” from Ancestry.com should indicate that.
The 1870 Census database at Ancestry.com indicates that Charley is in a household by himself.
It does show dwelling number 110, so that’s entered correctly.
But Ancestry.com’s automated “household system” should not separate these individuals out.
Those of us who always look at the original (which you should do) would easily see exactly who was in the household.
But this is one of those “automate” the research process “helps” that doesn’t.