What is horizontally partitioned data?
DataSHIELD is intended for the pooled analysis of ‘horizontally partitioned’ data, i.e. the collaborating organisations hold the same set of variables but on a different group of participants as depicted in the below chart.
To preserve privacy, can DataShield calculate for example X1 - X2 between partitions without revealing the values to the querying person/machine?
If so, can you explain how?
Thank you!