What your formula is doing is counting the number of blanks in the row in Table2 corresponding to the row in Table1. If the the data in the first columns of either table differ at all, you will not get the correct result.

The simplest solution would be to combine the tables. I understand your sample is likely a very simple example of your actual tables so it might be cumbersome. But as it is, both tables appear to have the same primary key i.e. the testees' names and thus theoretically would work.

There's certainly a way to do it the way you have it set up now, but I can't think of it. I'll ponder it.