How to Evaluate Training Programs
Kirkpatrick's Four Levels
Most (if not all) Instructional Design practitioners are familiar with the ADDIE model. While the evaluation phase (phase five) appears to occur last in the ADDIE model, in practical application it is an integral part of each of the preceding four phases. Evaluation occurs within each phase, between the phases, as well as after the implementation. It is important to build an evaluation plan into the design blueprint from the very beginning.
There are two types of evaluation, namely formative and summative. Formative evaluation occurs during and between the first four design phases, and its intended to improve the instructional materials prior to final implementation. Summative evaluation assesses the overall effectiveness or impact of the instruction.
Donald Kirkpatrick was the originator of one of the most widely used frameworks for evaluating training programs. Here is a summary of the four levels found in his work:
Level One: Reaction
How students react to the program. It is a measure of satisfaction from the consumer’s point of view. It does not necessarily correlate with learning. In fact, if mastering the training requires significant effort and concentration, the “smile sheets” as they are called, may reflect some measure of student dissatisfaction. This is true even if the training itself was effective.
Level Two: Learning
The extent to which students changed attitudes, improved knowledge, and/or increased skill as a direct result of the program. To eliminate external variables, this measurement is taken directly at the end of the program. Typically, some type of test (either knowledge-based or performance-based) is used.
Level Three: Behavior
The extent to which students change their behavior in a real world setting, because they attended the program. Students learn what to do and how to do it in the program. In the real world application, they must possess the desire to change their behavior, work in a climate that is conducive to the change, and be rewarded for the changed behavior. If the newly acquired knowledge, skill, or attitude is not beneficial to the student in the work setting, behavior will revert to pre-program levels, or transform into other behaviors that are reinforced and rewarded. The findings in a level three evaluation are used to improve program design.
Level Four: Results
The final results that occurred because the students attended the program. These results include, but are not limited to, increased production, improved quality, decreased costs, increased sales, reduced workforce turnover, and higher profits. If the objectives of the program are not stated in these tangible terms, then a level four evaluation will be nearly impossible to perform.
For example, if the objective states, “Students will gain an appreciation for sales forecasting,” a level four evaluation would be impossible to observe objectively. If on the other hand the objective states, “Using sales forecasting methods, students will be able to increase profits by 5% over their competition,” then objective data can be collected to confirm or dispute whether the objective was achieved.
All too often, training evaluation is either skipped entirely, or stopped short at level one. This is truly a shame. Even if only a portion of the training classes are sampled, so much can be gained by tracking behavior change and end results.
This content is accurate and true to the best of the author’s knowledge and is not meant to substitute for formal and individualized advice from a qualified professional.
© 2014 Carolyn Fields