Schools are changing, life is changing. So, any one person's experience might be quite different from another's, in school or at home or anywhere else. Many people of my generation experienced school, especially the years of 3rd grade and beyond, as a sampling and a surprise. The idea in such a set-up is that any test will be a sample of what the student knows, a sort of random check. If the student can answer 90% of the test questions, it is assumed that he knows 90% of the material that was studied. A test or "surprise quiz" in that way of verifying learning is not to be revealed to the students beforehand because of the danger of "studying for the test", that is, learning only answers to the test questions, spoiling the inference from test performance to all of the material studied.
I studied statistics in grad school and taught the subject in college. However, it was about 15 years after I began teaching, I looked into the work of W.E. Deming, Walter Shewhart and Joseph Juran. They worked in an area called "process control" and sought to make mostly manufacturing processes as efficient and error-free as possible. In their work, a car that comes off the assembly line that doesn't run properly is a big waste. They want as close to 100% of their creations to work as they can possibly get.
If we are really trying to get every student to learn, we too are trying to get as close to a 100% success rate as possible. There are political and social ramifications to high rates of success in education but teachers in tax-supported schools of a democracy try for success for all. So, just like the Scouts with their merit badge requirements, it makes sense to put the final right out there where students can keep an eye on the goals. Right from day one, they can benefit from knowing what they need to be able to do to show mastery. Some of the final might already be known before instruction begins.
Main blog: Fear, Fun and Filoz
Main web site: Kirbyvariety