Around 1967, the Coleman Report on educational performance compared the performance of schools and determined that the instructional program being used did not matter. The performance of students in schools as a whole appeared to be determined by the socioeconomic status of the students attending it. Instructional quality did not seem to matter.
Not long after that, educational researcher John Goodlad tried to determine which instructional programs in use at that time were the most effective. Unlike Coleman, though, he did not just compare total school performances. He went into the classrooms and watched teachers in action. What he discovered was that he could not possibly compare the effectiveness of different programs because many of the individual teachers were ignoring the programs they were supposed to be teaching and doing whatever they wanted.
Research since then has totally turned the Coleman study on its head and determined that the primary factor determining student success is individual teacher quality. When you look at total school performance, what you get is a conglomeration of great teacher performances, mediocre teacher performances, and poor teacher performances--all within the same school.
In a study a team I was on did on behalf of a large school district, an attendance area published the combined average results of a writing examination done at the 4th, 8th, and 10 grade levels. The average scores were roughly the same at all grade levels, and that is all the public was told. As researchers, we knew the hidden truth--not a single teacher had student performances anywhere close to that average. All the teachers either had student performances far above that average or far below that average.
That is what happens in public schools, where teachers have had years of intense and expensive training, and where teachers are evaluated regularly by professionals. Why would we expect scuba instruction to do better?