Re: st: Post codes to locality

[Gordon Hughes <G.A.Hughes@ed.ac.uk>]

To clarify the point that is implicit in other advice:
A.  This is a warning message not an error.  The situation in which 
one dataset contains multiple cases with the same merge id (postcode) 
is routine.  It gives rise to what is called "one-to-many" (1:m or 
m:1) matches - i.e. the same LSOA (and other data) is associated with 
each case with the same postcode.
B.  There would be a problem if you get the warning message that both 
datasets contain multiple occurrences of the same merge id.  This 
leads to a "many-to-many" (m:m) match: the results are predictable 
but probably not what you intended.  The details are given in the manual.
C.  The merge syntax introduced in Version 11 makes this explicit - 
i.e. either -merge 1:m ...- or -merge m:1 ...- specify a 
"one-to-many" merge.  Many of us are used to the older syntax which 
does not include <1:m> or the like, but -merge- will generate the 
same results provided that you are not using a m:m merge. You may 
think of the new syntax as way of checking that you understand your 
data and will get what you want.

Gordon Hughes
g.a.hughes@ed.ac.uk  

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/