No project requirements
were given to the control group. All three courses were taught by the
same instructor, and the study minimized the possibility of data manipulation
in the experimental groups by announcing at the start of the course that project
success would not affect their grades.