As I see it, you are in a bit of trouble!
If a given record is duplicated exactly, that doesn't present so much trouble, however, trying to only delete records based on duplicate SSN's is going to run into some data integrity issues, such as: how are you going to determine whether or not there is any other data that IS different, and if so, which record of the two duplicates should then be kept?
First of all you need to lay out this scenario in your own mind, then check the existing data to see if there are, in fact, such records. Then decide with supervisors, etc. what to do about it!
Afterwards, the process of doing the delete will look something like this:
Always work on a COPY of the real tables until you are sure that your process will work as you want it to!
Create a SELECT query that pulls the duplicates. (Use the Distinct keyword for this). Change the SELECT query to a Make Table query (right click on the query in design view and choose Query Type>Make Table). Create a separate, temporary table to hold these data by running your Make Table query. Open your original table, delete all entries. Copy and paste into the original table from the temporary.
Regards,
Scott