At the beginning of the month we shared with you an article by Jim Kirkpatrick about your "training defaults". He defines these as hard-wired behaviors that have become habits, but may not lead to the highest level of results. These are things we do automatically without considering if they are outdated or less effective than other options.

This week we encourage you to think about a common Level 2 "default" many training professionals possess: automatic use of pre and posttests.

There is certainly a time and a place for pre and posttests, but it is not before and after every learning event. Please "uncheck" automatic pre and posttests as part of your Level 2 assessment methods.

Testing is often time-consuming, expensive, and traumatic for training participants. Be purposeful about the use of pre and posttests. Situations where it is recommended include:

• Prepositioning the learner to the material to be covered in training

• Determining which elements of a course to emphasize and/or deemphasize

• Allowing employees to earn the right to test out of a program

• Demonstrating the amount of knowledge acquired during a formal learning event (if value of the event itself is being questioned)

• Meeting compliance issues

• Ensuring that participants are ready for more advanced training

If you decide to use pre and posttests, be realistic about the value of information they can provide. This form of assessment generally taps into the lowest levels of Bloom's learning taxonomy in the cognitive domain. This kind of data (e.g., "Our participants increased in knowledge an average of 37%") typically elicits a response of "Who cares?" from stakeholders.

To maximize training evaluation resources, and ultimately training effectiveness, focus your energies on assessing Level 2 skill demonstrations. Use the resulting data as part of your chain of evidence to show how learners were prepared to perform the required behaviors on the job. Redeploy the resources you saved by skipping pre and posttests to on-the-job reinforcement of the critical behaviors. This will drive the results the training was likely designed to deliver in the first place.