Education in scientific fields guided through non-scientific guesswork
I was having a rather interesting discussion with a friend of mine about the term Computer Science. Near the middle of this discussion he mentioned that, "There is no science in Computer Science." I happen to agree, if the conversational definition of the term 'Computer Science' were closer to the term 'Software Engineering'. Many conclusions and advances in the field of Software Engineering are not scientifically based. Among the many reasons why they are not scientifically based: they are too hard to reliably repeat, or it's too hard to get a sample similar enough when you repeat the experiment.
The same should be said of education in scientific fields. There is no science in the education of science, or, there is little science in the education of science. I recognize that there is some science involved, some application of the scientific method to reach conclusions - but not enough. If the scientific method were applied more liberally across the evaluation of education in scientific fields we'd be better able to grasp solutions to problems that currently evade our understanding.
Of the several problems that exist in the process of educating someone in a scientific endeavor, declining interest of women would be a key example. Since the scientific method is so little applied, we are not able to answer fundamental questions on that topic. One of the fundamental questions we can not answer is, how do we encourage more women to graduate with an advanced degree in Computer Science? Another could be, how can we encourage women to be interested in the possible applications of the scientific method?
One suggestion: apply it yourself. Take the scientific method and apply it to solve the problem. It will take time. It will experience growing pains like any other field, but in the end you'll be better equipped to surmise more than a guess. You might actually be able to solve the problem. Your actions might finally be effective.