Magazine article Training & Development Journal

An Integrated Evaluation Model for HRD

Magazine article Training & Development Journal

An Integrated Evaluation Model for HRD

Article excerpt

An Integrated Evaluation Model for HRD

Imagine this scenario: An HRD director is sitting in her office when the telephone rings. It is the chairman of the board, who says, "Say, some of the other directors and I were just sitting around, and one of them asked if all the fancy HRD programs we have around here actually do something and whether it makes any difference. Seemed like an interesting question, so we decided to ask you to sort of drop by, say sometime in the next 10 minutes, and just bring along any data you have that shows what we get out of our HRD effort. We're up in. . .Hellow? Hello?. . .Are you there?. . .Hello?"

This fictitious scene highlights what I think is a common and unfortunate condition in HRD today: the HRD manager up the creek without an all-in-one model for measuring program quality.

That's a tall order, but such a model does exist. It's an evaluation system that integrates with needs analysis, design, and delivery. Armed with information produced by this model, our heroine could have raced gleefully up to the chairman's office with an armload of convincing data. More to the point, had she been using this model, the chairman never would have called in the first place, for the HRD director would have asked and answered many times over the chairman's very good question.

Yesterday's evaluation system

There is no doubt that the HRD profession is enjoying the best of times--never before have organizations invested so heavily in our services. And in such good times, it can be difficult to think of evaluation and accountability. Most of us are too busy serving the next client's demands to bother to see whether the last client benefitted. But as the HRD profession does more, promises more, and inevitably requires more resources to support itself, we need--now more than ever--comprehensive and effective evaluation approaches.

In 1967 Donald L. Kirkpatrick proposed a four-step model that has, in many respects, provided a sound and simply understood conceptual base for evaluating HRD programs. Appearing as part of the Training and Development Handbook, edited by R.L. Craig and L.R. Bittel, Kirkpatrick's model very clearly articulates four levels of outcome for any training session. Each level, he wrote, demands separate evaluation: Did trainees react favorably to the training--did they like it? Did trainees learn? Did trainees use what they learned? Did using the learning make a difference? This definition of four levels of outcome is extremely useful, for it pushes the focus of evaluation beyond mere favorable reactions and learning to where it rightfully belongs: on payoff to the organization.

Yet I find important shortcomings with the Kirkpatrick model. First, its definition of HRD covers training only, programs in which trainees learn discrete skills that transfer readily to the workplace, producing immediate results. Today's HRD efforts are much broader than that.

Consider, for example, applying the four-step model to a program teaching people how to save lives by using CPR techniques. Barring a workplace heart attack, we would find no on-the-job application of the skill learned. Does this failure to detect broad third-level effects negate the value of the program? Probably not, so we need a different way to evaluate this program properly. Some HRD programs that do not produce behavioral results may nonetheless have value. Others may deserve to be thrown out. We need an evaluation process to help us decide not only which programs to keep and which to discard, but also how to revise those we keep to make them more cost effective. The Kirkpatrick model's narrow focus prevents us from doing this.

The Kirkpatrick model is entirely outcome oriented, reflecting a legitimate bottom-line bias. Yet there are many reasons to be concerned with evaluating HRD programs as they happen, well before they have had a chance to produce results. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.