One way to measure trial$-$and$-$error learning in the laboratory is to use a:

nonmatching$-$to$-$sample task.

concurrent discrimination task.

displacement task.

competitive objects task.