Abstract
Programming education has experienced a shift from imperative and procedural programming to object-orientation. This shift has been motivated by educators' desire to please the information technology industry and potential students; it is not motivated by research either in psychology of programming or in computer science education. There are practically no results that would indicate that such a shift is desirable, needed in the first place, or even effective for learning programming. Moreover, there has been an implicit assumption that classic results on imperative and procedural programming education and learning apply to object-oriented programming (OOP) as well. We argue that this is not the case and call for systematic research into the fundamental cognitive and educational issues in learning and teaching OOP. We also present a research agenda intended to improve the understanding of OOP and OOP education.