In 1992, the economist, Gary Becker, won the Nobel prize for his work that demonstrated the importance to organizations of human capital and of training, in particular. Despite the importance of human capital to the long-term health and growth of organizations, they continue to under-invest in training (Becker, 1993). In The Human Equation, Jeffrey Pfeffer (1998) explained why, “Training is an investment in the organization’s staff, and in the current business milieu, it virtually begs for some sort of return-on investment calculations” (p.89). In other words, because organizations do not adequately measure the value that training adds, they fail to reap the benefits of fully investing in training. This article shows you how to measure your return on investment in training.
Main Focus: How To Measure Your Training
You can follow a simple four-step process to achieve valid measurement of your training efforts:
Key Terms in this Chapter
Statistical Significance: A measure of how confident you are that the result you got did not happen by chance alone. It is not an indicator of how meaningful the result you found is. With enough data you can be very confident that you found something very unimportant.
Opportunity Cost: The next best alternative you could have had with your time or money. The opportunity cost of a salesperson taking sales training is sales the person could have made while attending the training.
Psychometric: Having to do with the branch of statistics that relates to measuring the psychological aspects of people, such as personality traits, intelligence, knowledge, and skills.
Theory: A system of ideas that explains a phenomenon by connecting empirical data to other data and ideas (Waddington, 2007).
ROI / Return on Investment: Defined as (Value-Cost)/Cost (Waddington, Aaron, & Sheldrick, 2005). For example, an ROI of 22% tells you that for every dollar spent, you get $1.22 back.
Correlation: A statistical measure that varies between -1 and +1 that indicates the degree two which two things are related. A correlation only indicates that the two variables tend to move together; it does not necessarily mean that one variable causes the other.
Regression: A statistical technique that allows you to determine the mathematical relationship between one or more independent variables and a dependent variable. For example, given that car accidents increase in bad weather and on weekends and that bad weather tends to be more common on weekends, regression allows you to isolate the effect of accidents due to weekends alone.
Validity: The degree to which both theory and empirical evidence support the intended actions and inferences based on the data. It is not a property of the data, but of inferences made from the data (Messick, 1989).