<p>Problem sets’ construction poses one issue: the writer’s views on what’s important may not be in line with the worker’s, and legitimately these two may just care about different things (in depth). There’s usually some material that is streamlined and needs to be done a certain way, and problems help one work with the ideas. I think most would say this is even true in research – one can read through masses of theory, but actually working with the ideas will tend to lead one to really have some direction in how to wade through the theory, even beyond their own scope.</p>
<p>Overall, I think a firehose system is most effective if it offers choices. For instance, taking courses on a pass/fail basis can take the pressure off the individual to obsess over every last detail and focus on what’s important to the individual. But having the option to take classes that really challenge one to the limits from a pure problem solving standpoint can have its benefits too. Certainly there should be a balance. Unless one is sure that firehosing works well for individual taste of course!</p>