Hash Example.

Example hash of first name, last name and date of birth. The last 7 digits of the SHA-256 hash value are kept and the rest of the hash discarded because of memory limitations. The 7 digit hash is a numeric hexadecimal value. For step 1, each round of the Yahtzee procedure begins with the hashing of the datasets using a new salt. The “salt” allows us to generate multiple hashes without getting the same hash every round. Next, the hash is divided by the value , where is the number of individuals in the dataset and was chosen arbitrarily. The remainder of this calculation is recorded as the group ID. Records are then placed into groups of various sizes based on this group ID. On average the groups should contain records. Next the frequency of some behavior of interest - in our case voting - is recorded for each group ID. In subsequent steps, a group ID is generated using the identical process on a second dataset. In the second dataset, the frequency of the behavior of interest is assigned to each record based on its group ID. In some cases, the same record is in both datasets, and its contribution to the value assigned to the group in the origin dataset will be transferred to the group in the destination dataset. However, individual records are never matched. We can be sure that identical records in both datasets will be assigned the same group ID, but we can never be sure for any one record if a true match exists in the other dataset or just records that hash to values with the same remainder after dividing by .