Hello stata users,

I am working on a propensity score analysis and using
Code:
pscore
to create it. One of the options there is
Code:
blockid
which is defined as "specifies the variable name for the block number of the estimated propensity score" by Becker and Ichino 2002.

I don't quite understand what it does. What's the optimal number of blocks for a dataset, if any? What happens when you exceed a certain threshold of blocks? Please guide me to appropriate resources, if possible.

Thanks.