Continuing from For Example: “Experimental design principles for Ottawa–Toronto
f lights,” you’re considering splitting up the students into two groups before mailing
the offers: those who live or go to school in Montreal, and those who live or go
to school in Ottawa or Toronto. Using home and school postal codes, you split the
original 16,000 students into those groups and find that they split 8000 in Ottawa or
Toronto and 8000 in Montreal. You plan to randomize the treatments within those
two groups, and you’ll monitor them to see if this factor, City, affects their Miles Travelled
as well as the type of offer they receive.

QUESTION 

What kind of design would this be? Diagram the experiment.