Controlled experimentation is therefore the evaluation process of choice. A controlled experiment uses both a training group and a control group that receives no training. Data (for instance on quantity of sales or quality of service) are obtained both before and after the group is exposed to training and before and after corresponding work period in the control group. This makes it possible to determine the extent to which any change in the training group’s performance resulted from the training, rather than from some organization wide change like a raise in pay. (The pay raise should have affected employees in both groups equally.)
Controlled Experimentation is a formal methods for testing the effectiveness of a training program, preferably with before-and-after tests and a control group.