Steve, that approach has a lot of problems associated with it, and in my opinion, it creates “data” which appears useful, but is potentially misleading. It’s also a quite typical “government response”, and I don’t mean that in a negative way exactly, but just in the sense that the process is a little overly-complex.
The only context I would advise doing something like that is if it’s a really large scale needs assessment, let’s say across more than two hundred employees.
The reasons? First, what you describe means a LOSS of find grained data, and person to person interactions, and the major flaw is the ratings system, which causes data loss, and ends up encouraging a paper chase (or e-chase). Also, while you might think you can add ratings, and operate on them mathematically, you can’t, or at least not in any meaningful way.
I haven’t written anything specific to rating scales as a part of training needs assessment, but I have covered the problems if anyone is interested, particularly within the context of managing performance.
I do understand that what you describe has become more “standard”, particularly because one can use computers and software to do a lot of the work, but sadly, it’s often garbage in, garbage out.