The pretests that they show me are usually long, complicated, and intimidating forms asking the testee very specific knowledge points that all the participants specifically came to the training to learn. The post-test is almost always the EXACT same test as the pre-test; the only difference between the pre-test and the post-test being the headings PRE-TEST and POST-TEST boldly typed in imposing font at the tops of each test.
I will usually counter with my arsenal of brain research about how people best learn new information. I talk about how we do not wish to make the learners feel threatened in the very first hour by prompting them with big scary tests asking them about information that we already know they don't know. Pretests usually result in learners feeling threatened or embarrassed or frustrated or even fearful because by administering pre-tests, we imply that somehow we expect learners to know information that they don't yet know because we have not yet taught them. I explain that when learners feel threatened or embarrassed or frustrated or even fearful, their fear/rage response is triggered and well, all learning simply comes to a screeching halt. And then when the trainer or teacher collects the pre-tests and starts to teach, they have already diminished the group's ability to learn significantly. Not a great way to start a teaching or training event.
And yet, when those same trainers and teachers administer that same test - this time with the POST TEST heading - they can show significant gains in participant knowledge. Well yeah, that's because the trainers/teachers actually taught the information to the participants. No big surprise that there will be gains when you compare results. I believe strongly that if you don't administer that intimidating pre-test in its present scary form and test for prior knowledge in a more authentic way, those post-test results will be even greater.