To evaluate a program designed to improve math scores for elementary school girls, girls were divided into treatment and control groups. Girls in the treatment group were required to attend an after-school program, while girls in the control group went home on time. Math exam scores were then compared between the two groups of girls. This type of program design is associated with: