Yes, in this case I can. But we have many, many, many tables and I don't know the number of records of any table by heart. So any time I want to know if it's a partial scan I have to determine the full number of rows first?

The point is, many of our foreign keys reference to columns that are the clustered index so many joins result in clustered index lookups or scans. And that's where I want to know the difference. As the scan is done maybe multiple times I cannot even have a look at the logical reads as they are multiplied.