I've been writing about a death certificate and obituary from 1848 on my Rootdig blog recently.
Essentially the death certificate and obituary agree on everything--except the state in which the person died. The obituary says Kingston, NJ and the death certificate says Kingston, NY. One might be tempted to say the death certificate is correct and the obituary is not and that "NJ" can easily be a typographical error for "NY." This is true.
However, it's not clear just how the 1848 City of Philadelphia death certificates were created.
It is also worth noting the proximity of all the places involved--how plausible is it that New York State is correct in this case?
The 1848 death certificate is discussed here.
The 1848 obituary is discussed here.
Looks like an interesting discussion of more than just primary and secondary is in order. The nature of the record, learning how it was actually created, etc. all play into this analysis.