Matching using reclink

[Rajdeep Chaudhuri]

Hi,
I have two datasets, one at the village level and the other at the school level. Now the village names across these datasets are different in spellings, leading me to assume that fuzzy matching is the way to go about it if I want to merge on the village names. However, is it possible to use reclink to do this type of a fuzzy match, since each village name would be repeated more than once in the school level file(as each village can have more than one school). ?
In other words, I am asking whether a m:1 (or 1:m) type merge can be done using reclink, or is there any other way to go about it?

[Clyde Schechter]

-matchit- will pair up each observation from the village data set with as many observation in the school data set have a village name that is a plausible match to the name in the village data set. If it turns out there is only one such, you will get an m:1 match for that village. If there are many, you will get many:many matching. (I am deliberately not calling that m:m to avoid confusion with the dreadful mess that -merge m:m- creates and which should never be used.) If that village appears only once in the village data set you will get 1:1 or 1:m as the case may be.

Otherwise put, the arity of the match depends on the availability of suitable matches in the data sets; it is not predetermined in the code. That said, in situations such as the one you describe, you are likely to end up mostly with 1 village matched to many schools.