I have encountered a couple of those cases, and I agree with the others that there are several possibilities, and you should not leap to a conclusion based on this data alone. One case was very similar to yours, where an entire family had been counted twice. Although there were some differences in ages and other data, I reached that conclusion because there were other neighboring families that were also listed in both records, and I could find no other records to suggest that another family with the same or similar names had ever existed in the area. It appeared that two different census enumerators had just inadvertently traveled down the same country road a couple of weeks apart, and the residents really were counted twice.
In your case those two records appear to have been created by the same census enumerator in the same locale, and (with a quick look) I didn't see any other duplicate entries nearby. So yes, maybe there was one family that moved, or perhaps there were two different families. If it's the latter, then there may be other records that could help to sort it out -- e.g., birth or death records, marriage records for the children, later census entries after the children reached adulthood, etc. If you don't have any conclusive evidence at this point, I agree with Nan that you should include info about both records in the bio, and let any future researchers know what you have found and why there's uncertainty.
