!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> Streamline Training & Documentation: The Kirkpatrick Model at 50

Thursday, November 12, 2009

The Kirkpatrick Model at 50

It was in 1959 that Donald Kirkpatrick published the first of four articles in the Journal of the ASTD elucidating his four-level model for evaluation of training.1 Krikpatrick had initially defined the model — though not with the levels' current names — Reaction, Learning, Behavior, Results — in his doctoral dissertation, "Evaluating a Human Relations Training Program for Supervisors."

Fifty years have elapsed, and the four-level model has held up well. The anniversary is being marked by various review articles and interviews with Kirkpatrick himself and with his son, Jim, who has been carrying on in his father's line of work for some years.

Jim Kirkpatrick, along with his wife and co-author Wendy, has provided a fresh overview of the model in a white paper (reg req), the highlight of which is the diagram below. The diagram shows the Kirkpatricks' take on the flow of activities involved in developing and evaluating a training program.

(click to enlarge)

(from a white paper by Jim & Wendy Kirkpatrick)

ROE = Return on Expectations,
i.e., the degree to which the training's results meet management's expectations concerning what the training should accomplish.

KSAs = Knowledge, Skills, Attitudes,
i.e., competencies.

The bulk of the ten-page white paper is devoted to explaining how the Kirkpatrick Model embeds the four levels into the planning of the training, as well as its evaluation. There is considerable attention to the practical details of the process, which makes reading the white paper in its entirety worthwhile.

As a quick summary, here is an example of the process in practice that Jim Kirkpatrick lays out in the interview with him and his father that Training Magazine published in October:
Let's say an executive in your organization asks for a new program on leadership. Instead of conducting a cursory needs assessment, and putting together and delivering a program, start with the concept we call "the end is the beginning." You must begin with a clear picture of what a successful leadership program ultimately will bring to the organization in terms of Level 4. This means you first negotiate your stakeholders' expectations so they are satisfying to them and achievable for you, then you convert those expectations to observable, measurable success outcomes by using various forms of the question, "What will success look like?" (Level 4, Results). Then you determine the critical new leadership behaviors training grads will have to perform on the job (Level 3, Behavior) to bring about those success outcomes.

Also crucial at this stage is to determine, implement, and monitor the required drivers of those behaviors — the processes and systems that will serve to get grads to perform those critical behaviors on the job. Next, it is important to identify and reinforce pre-training necessities for success, for example, a culture of support and accountability, coaching guides, and job aids. Then, and only then, you develop and deliver the training that will produce competent employees (Level 2, Learning, and Level 1, Reaction) who will be able to perform the targeted, critical behaviors. In other words, start with Level 4 and work backward.

It is also important to monitor not only Levels 1 and 2, but the required drivers and critical behaviors at Level 3, and the outcomes at Level 4. This will allow for a good system to make sure learning to behavior transfer occurs, and also provide you with a compelling chain of evidence at the end to demonstrate the impact of the initiative.
As you can see, there is nothing obscure about the process. Using the four levels effectively is a matter, as the Kirkpatricks emphasize, of strong partnership between training professionals and organization management, accompanied by excellence in implementation.

1 Donald L. Kirkpatrick, "Techniques for Evaluating Training Programs," Parts 1-4, Journal of ASTD, Vol. 13, No. 11 (1959), pp. 3-9; Vol. 13, No. 12 (1959), pp. 21-26; Vol. 14, No. 12 (1960), pp. 13-18; Vol. 14, No. 13 (1960), pp. 28-32.