Like anyone who had much experience with U. S. Census records, I've seen a lot of errors, but this is a new one on me.
The death of Elbridge H. Beckler in August 1908 is clearly documented, both by an obituary published the next month, and by a FindAGrave memorial. But there he is in the 1910 Census, living at Chicago as a widower with his two daughters, with no wife (who was very much alive) shown. The age matches within a year of his wife's age at the time and is off by five years of what his would have been.
My guess is that the information came from one of the daughters and was misunderstood by the census taker, who thought it was the mother who was deceased and not the father.