Science is often portrayed as an activity that provides answers. And so it does.
To get to the moon, you really do need a certain amount of thrust. There really is a weight limit to this bridge. The cannon you just shot at the top of a mountain really will fall to earth along a predictable curve.
Perhaps by analogy to the natural sciences or perhaps simply from a belief that there must be a correct answer (coupled with a desperation to find it), many people seem to expect the same kinds of answers from sciences of human behavior. What is the best learning method? What is the policy that results in the best outcomes? What app should I use?
When I answer questions like these about learning, I find myself using the words, "generally speaking," a lot. Often, these questions are framed in a way that's too general to be helpful, but there's usually some basic, abstract things to say. But, "in general," the research in psychology, policy-making, economics, anthropology, cognitive science, and many other disciplines focused on human behavior doesn't provide the same kind of value that, say, orbital mechanics provides for rocket engineers.
The difference is easiest to discuss by reference to the "state-space" interpretation of problem solving. Problems can be in different states, and we move through different states to solve a problem, looking at possible alternatives, weighing the costs and benefits, and picking the next "best move". Like a chess problem.
If science has fully "solved the problem," then following the scientific principles and mathematical equations provides a path to the goal state. It tells us what we are going to do at every state. For nearly every adult, tic-tac-toe is a solved problem: we know what to do at every state of the game to avoid a loss.
But in ill-structured domains, like learning and teaching, the research provides value by helping us to figure out what state we are in, what the possible alternatives are, and the tradeoffs of those alternatives. The same is true in policy-making, in economics, and in various other applied social science fields. It's not that the research tells us exactly what to do. It's that the research informs us about what we're doing and what we should do next.