It is important to understand the importance of ensuring your measures of training effectiveness are reliable and valid. Any metric used by organizations must meet these criteria to be useful. Consistent with how we assess job candidates, a measure is only useful if it is predictable and measures what it is intended to measure. A reliable measure must be consistent and measure the same thing over time, across different evaluators or internally in an instrument.

This applies to training because we want to be sure each time we assess a specific training outcome—satisfaction with the trainer, for example, or assessing learning post-training—it gives us consistent results. In the context of needs analysis, if we want to assess trainees’ skills and knowledge accurately in a pretest measure, we also need to be sure the measure is consistent.

A valid training measure must assess what it says it will assess. For example, if I argue a training measure assesses learning, I need to ensure that it actually measures learning, not some other aspect associated with the training process. Similarly, a valid training measure assesses the ability of the measure to accurately reflect training outcomes.