In the past few decades, an increasing number of physicists have taken up the challenge of applying methods of research based on those that have been employed successfully in investigations of the physical world. This endeavor is broadly known as "physics education research" (PER). Systematic studies of student learning have revealed a wide gap between the objectives of most physics instructors engaged in traditional forms of instruction and the actual level of conceptual understanding attained by most of their students. But PER has gone beyond documenting shortcomings in student learning and traditional instruction. Researchers have developed instructional materials and methods that have been subjected to repeated testing, evaluation, and redesign. Numerous reports have documented significant and reproducible learning gains from the use of these materials and methods in courses ranging from large-enrollment classes at major public universities to small classes in two-year colleges and high schools. Still, there remain inadequacies in even the most recent instructional approaches and many unanswered questions.

In this Guest Editorial we identify some of the current and emerging research directions that we consider promising. We also argue for the importance of doing research on the learning and teaching of physics in physics departments. We do not mean to suggest that PER should not be conducted in schools of education, but, as we argue later, we do not believe that the field is viable without a critical mass of faculty in physics departments. Finally, we identify some practical and political challenges and propose some steps that could be taken to help ensure the stability, growth, and productivity of PER.