Just continuing to look at the different ways people measure training. Measuring behavioral change isn’t a bad one. If you really know what to do look for, you should start to see changes being made after training. Often creating behavioral checklists and doing direct observation is a popular way of measuring at level 3.
Here are the pitfalls. First, when you divide things into behaviors you can loose how they work together. You can do all the new behaviors but miss all the connections. More than likely, participants begin to get better at these behaviors but haven’t yet reached a level of proficiency or mastery. That takes a lot of time, practice and feedback.
Look at the example of learning to do great presentations. You can train to a set of new behaviors in front of others during a presentation. You can then look to see if those behaviors are starting to appear. You should see some change. This is a good thing. However, to continue on to a higher level of proficiency won’t happen immediately. In fact, the new behaviors can quickly dissappear under the pressure of doing things for real.
The solution is to look at how you can continue to build those behaviors and work on all the subtleties that may not be on the checklist.