To establish causation you’d have to randomly assign people into groups, not let them self-select into marijuana users and control.
To establish causation you’d have to randomly assign people into groups, not let them self-select into marijuana users and control.