I read an analysis of two 19th century US census entries that contained different information for the same immigrant family. The reason for the difference, according to the analysis, was that the informants were probably different.
That’s a very definite possibility and if the respondent in one entry spoke better English than another respondent did the answers may have reflected that.
But it’s not the only reason that the enumerations could have been slightly different. There was another individual involved in that census enumeration who could easily have impacted the accuracy of the entry:
- Was the enumerator able to understand the response of person giving the information?
- Was the numerator familiar with the language the respondents probably spoke?
- Was the enumerator familiar with the foreign-language names of the family?
The informant is extremely important–when known. But the analysis isn’t just about the informant. It’s also about the person hearing that information and writing it down.