A check digit is a way of
validating the
integrity of your
data (see
data integrity). By using an
algorithm to create a digit unique to the
data being
transmitted you can then use the same
algorithm on the other end (or at a different time) to check whether the
data is
valid. If it has got
corrupted the check digit will probably be wrong for the number. This check digit is usually tagged onto the end of the data being
transferred.
An example where a check digit is used is for ISBN numbers where the Mod 11 hash function is used to create a check digit.
lj had this to say on the matter of a check digit which I don't understand totally so I'll just copy and paste and hope that this has more meaning for others.
'Because of the pigeonhole principle ("if you have n holes, and n+1 pigeons, at least one pigeon has to share"), a check digit can never be unique to the data it hashes, as it contains less entropy than the hashed data.