Responses inline:
You don't know what the case or the scope of the questions will be for a
given solution.
I have worked on insurance applications that asked the insurance agent or
broker over 900 questions in various categories, with all of it being
displayed on one ASP.net page by selecting the various categories from
combo boxes with any where from 1 to 50 questions at a time with the
answers to questions given in combo, textbox and radio controls in a table
or grid, and all of it being built on the fly.
It sounds like you provided an excellent solution for their specific
problem. You are to be commended.
I have also done the same thing with questionnaire surveys for car
companies and their dealerships with their customers when asking about
features in cars based on make and model.
You apparently did a good thing, yet again.
You can pass 10,000 MS certification tests with screen by screen
questions. That doesn't make it a standard that should be followed.
I never stated or implied that what I was recommending was a standard. The
point I was making was that for some tests, it is perfectly reasonable to
present one question at a time. By bringing up the MS certification exams, I
was simply illustrating a real-world example of where this might be the best
approach. By stating the number of exams, I was pointing out that my
observations and experience are not entirely unique to one special case.
There was this one programmer who did just that. He developed a solution
for some in-house users that went screen by screen or page by page asking
questions, they wouldn't use his application, and they talked about him
like a dog.
So apparently he did the wrong thing. But that doesn't mean that presenting
one question at a time is always the wrong thing. Apparently you are
unfamiliar with progressive tests whereby expertise in some subject matter
is being measured via some "progressively determined" or dynamic set of
questions... whereby answers to early questions determine which questions
are presented next. Where expertise is confirmed early in a test, future
questions on that subject domain are omitted. Where expertise is not clearly
demonstrated, additional questions within the subject domain are dynamically
added to the exam... all this in an effort to get a more accurate
understanding of the expertise being measured. There is a lot of theory that
goes into these types of tests. You can google "psychometrics" for more.
Furthermore, what about plain ole' wizards. Sometimes they will present a
single question that determines what successive wizard pages are presented.
You don't know the scope of the of this, what I have suggest my be the
best approach, and the OP doesn't know that or doesn't know any better,
not to go screen by screen.
You don't know enough about the OP's motivations either, and jumped to a
conclusion that the OP doesn't know any better. I was simply pointing out
that you are wrong about that, provided some real-world examples (at least
19 that I have firsthand knowledge of).
My intention is to not be confrontational - but where somebody (the OP in
this case) is given bad advice, I'll point that out.... as doing so helps to
maintain the value of these NGs.
One programs the screen by screen solution may be a viable solution, but
on the other hand, it may not be a viable solution too.
Great! Apparently we agree. It's too bad that you didn't consider that with
your response to the OP and instead had to imply that the OP was doing
something wrong and proceeded to give very explicit advice to do something
based on your assumptions. To quote you: "You need to hook up a datagrid
with horizontal and vertical bars to move up and down the questions or left
or right on a question...". There is nothing there, or elsewhere in your
post, that helps the OP to solve their original problem.
-RC
