I've never been to technical training that required a test after completion. There is always the post training questionnaire to gauge things like the instructor's presentation skills and the instructor's ability to field questions. In theory, I could show up drunk to training, learn nothing, and still receive my certificate; which, of course would certify that I am an expert in the given technical area ;-). My preference would be to have a pre-test and a post-test to measure how much I actually learned. To take this a bit further, I would prefer that the pre-test be given as a part of the training registration process. If you were to take the pre-test and fail, you would not be able to register for the training. Is any or all of this a good idea? Why or why not?
In general, employees do not like to be tested in a way that is tied to success or failure. So if you give an "exam" at the end of a class, you would probably see enrollment drop sharply. That ends up hurting the company, because you want your employees taking classes to continually improve their knowledge/skills.
The reason for this is pretty simple, I think. Employees are adults and want to be treated that way, and giving them an "exam" tells them their employers do not trust them to pay attention/be professional in a class. It sends a message that their employer thinks they are "children" who need to be "Watched" and creates stress for even good employees who become more concerned about what will be on a test than what they need to learn.
That said, I think pre- and post-tests to gauge what you have learned is good, so long as they aren't tied to successful completion of the class.
I think it would be a great idea to use pre- and post-tests to gauge the effectiveness of the instructor. Tests could be given to students anonymously, then the class average scores (pre- and post-) could be used to evaluate the instructor.
I'll add that training != certification, and the question seems to be confusing the two. Everyone expects to be tested before receiving a certificate indicating they've mastered something of any significance. Otherwise--such a certificate is beyond useless.
Testing is very helpful to the student and teacher. The teacher needs a feedback mechanism to see how successful they were. It would be nice if the student could see their results, but the teacher/management only sees the results as an aggregate. That way they can use the results to improve their program, and the student can use their own results to fairly assess how much they absorbed.
If the training program is critical to an employees job performance, ignore that and test away. A mastery of the content must be proven, or a retaking of the course (or other way of learning) redone.
I would not recommend testing, but instead let those taking technical training hold an internal training session for their co-worker where they focus on the highlights from the training they received themselves.
That may lead to them focusing more when they receive the training (not showing up drunk) and already there try to find good points they have to take back to their co-workers.
Another benefit is that the whole team learn something when one from the team is sent away for the training.
The above answers mostly assume a perfect world where everyone attends training sessions for identical reasons.
I've seen whole teams(programmers, testers, PM, TL, managers, and techwriters) sent to detailed technical training with the notion that everyone comes out with an idea of the effort and complexities involved with the technology AND with the capacity to all speak the same language or argot on the job. A method of arguable effectiveness to reach a goal to be sure.
I've seen people sent to technical training because there are significant discounts if you send 10 or more people to the class or because its the end of the budget year and what you don't use you lose for next year so you bundle everyone off to training.
And I've seen people sent to training because a VP somewhere decided that everyone must go.
In any of these scenarios testing will have a marked negative effect on the teacher and at least some of the students.