Any training or development implemented in an organization effort must be cost-effective.

The
benefits gained must outweigh the costs of the learning experience. It is not enough to merely assume
that any training an organization offers is effective; we must develop substantive data to determine
whether our training effort is achieving its goals. Did the training correct the deficiencies in skills,
knowledge, or attitudes we assessed as needing attention? Note, too, that training and development
programs are expensive—in the billions of dollars annually in the United States alone. The costs
incurred justify evaluating the effectiveness.
Evaluating Training
How will we determine if a training program is effective? This is easier if some output can be
measured such as an increase or decrease in costs, sales, production, employee turnover, or revenue.
In these cases, HR can calculate a return on the investment (ROI) by determining the benefit of the
training and dividing it by the training expense. For example, after training, a delivery driver is able to
make five additional deliveries each day. Each of those deliveries represents $10 of the driver’s time,
so the benefit to the employer is $50 per day for each driver trained. Divide the benefit of the training
by the cost of training, and we can determine the ROI of the training. What if we’re training managers
on better communication skills or teaching English as a second language to employees? That’s a little
more difficult to evaluate. A different approach to determining the effectiveness of training is called 
Kirkpatrick’s model.
This is a four-level approach that works well in determining the value of managerial training and any
training that is difficult to assess in terms of ROI.
 Level one 
measures the reactions of the participants toward the training and answers questions about whether the
participants liked the training; felt they achieved their learning goals; how much they liked
the trainers; and any suggestions they have for improving the training.
 Level two
measures how much the participants learned. This could be accomplished by pre- and post-testing the
participants or by evaluating the participants against a control group that has not been trained.
 Level three
measures whether the training actually changes the employee’s behavior when he or she
returns to the job. This might be evaluated by the participants, supervisors, or trainer.
 Level four 
measures whether the training benefited the employer or not. This could be done by determining ROI
as we have above, or by evaluating a behavior against another standard, such as a benchmark.
Surprisingly, research indicates that nearly half of all training programs are not measured against any
substantative outcome, such as employee retention, satisfaction, or productivity.

It would be ideal if all companies could boast the returns on investments in training that The
Cheesecake Factory does. Their employee turnover rate has been reduced to 15 percent below the
industry average; customer satisfaction rates and repeat visits are high.

Such a claim, however, is valueless unless training is properly evaluated. The following approach for
evaluating training programs can be generalized across organizations: Several managers,
representatives from HRM, and a group of workers who have recently completed a training program
are asked for their opinions. If the comments are generally positive, the program may receive a
favorable evaluation and it will continue until someone decides, for whatever reason, it should be
eliminated or replaced. The reactions of participants or managers, though easy to acquire, are the
least valid. Their opinions are heavily influenced by factors such as level of difficulty, entertainment
value, or the personality characteristics of the instructor, all of which may have little to do with the
training’s effectiveness. Trainees’ reactions to the training may, in fact, provide feedback on how
worthwhile the participants viewed the training.

Beyond general reactions, however, training must also be evaluated in terms of how much the
participants learned, how well they use their new skills on the job, positive changes in behavior, and
whether the training program achieved its desired results including reduced turnover, increased
customer service, etc.
Performance-Based Evaluation Measures
We’ll explore three popular methods of evaluating training programs. These are the post-training
performance method, the pre–post-training performance method, and the pre–post-training
performance with control group method.
Post-Training Performance Method
The first approach is the post-training performance method.
Participants’ performance is measured after attending a training program to determine if behavioral
changes have been made. For example, assume we provide a week-long seminar for HRM
recruiters on structured interviewing techniques. We follow up one month later with each
participant to see if, in fact, attendees use the techniques addressed in the program and how. If
changes did occur, we may attribute them to the training, but we cannot emphatically state that the
change in behavior is directly related to the training. Other factors, such as reading a current HRM
journal or attending a local Society of Human Resource Management presentation, may have also
influenced the change. Accordingly, the post-training performance method may overstate training
benefits.
Pre–Post-Training Performance Method
In the pre–post-training performance method, each participant is evaluated prior to training and rated
on actual job performance. After instruction—of which the evaluator has been kept unaware—is
completed, the employee is reevaluated. As with the post-training performance method,
the increase is assumed to be attributable to the instruction. However, in contrast to the post-training
performance method, the pre–post-training performance method deals directly with job behavior.
Pre–Post-Training Performance with Control Group Method
The most sophisticated evaluative approach is the pre–post-training performance with control group
method.
Two groups are established and evaluated on actual job performance. Members of the control group
work on the job but do not undergo instruction; the experimental group does receive instruction.
At the conclusion of training, the two groups are re-evaluated. If the training is really effective, the
experimental group’s performance will not only have improved but will be substantially better than
the control group. This approach attempts to correct for factors other than the instruction program that
influence job performance. Of the numerous methods for evaluating training and development
programs, these three appear to be the most widely recognized. Furthermore, the latter two methods
are preferred because they provide a stronger measure of behavioral change directly attributable to the
training effort