THE USE OF SELF-REPORT DATA

Self-report data, or data provided directly by individuals involved in the training program, is the most commonly used type of data in HR evalua- tion.78 Recall that trainee reactions (Kirkpatrick’s first level) remain the most widely used evaluation measure. Self-reports can offer personality data, attitudes, and perceptions and can provide information to measure the effectiveness of HRD or other programs.For example, the trainer may measure learning by asking trainees to judge how much they knew before training and how much they feel they know after training. Information col- lected this way, whether through interviews or questionnaires, can be useful and meaningful. However, two serious problems can occur when relying on self-report data:

1. Mono-method bias. If both reports in a before-and-after evaluation come from the same person at the same time (say, after training), con- clusions may be questionable. The respondents may be more concerned about being consistent in their answers than about providing accurate responses.

2. Socially desirable responses. Respondents may report what they think the researcher (or boss) wants to hear, rather than the truth. For example, employees may be fearful or embarrassed to admit that they learned nothing in a training program.

In addition, there can be what is referred to as a response shift bias, in which respondents’ perspectives of their skills before training change during the training program and affect their after-training assessment.81 For example, trainees may discover during training that their pretraining judgment of skill was unrealistically high and then adjust their post-training evaluations accordingly. As a result, the data may show no improvement of skill after training, even though an improve- ment may have occurred.