Another way to use OVER is to have a result column in your select operate on another "partition", so to say.
This:
SELECT
name,
ssn,
case
when ( count(*) over (partition by ssn) ) > 1
then 1
else 0
end AS hasDuplicateSsn
FROM table;
returns 1 in hasDuplicateSsn for each row whose ssn is shared by another row. Great for making "tags" for data for different error reports and such.