Innovation in Language Learning

Influential educational psychologist Bemjamin Bloom identified, as early as 1984, the “2 sigma problem”: namely that the average student tutored one-to-one performed two standard deviations better than those students taught in classroom settings. While Bloom acknowledged that one-to-one tutoring for all students was not practically possible for educational systems, he challenged educators to “find methods of group instruction as effective as one-to-one tutoring”.  This presentation will argue that the rise of artificial intelligence technologies and specifically intelligent tutoring systems - for the first time since Bloom first identified the 2 sigma problem – provide a practical means of personalized and customized instruction which may address the limitations of a group classroom environment. The presentation will begin with an explanation of the research underlying Bloom’s 2 sigma problem, including a description of the mastery learning techniques which are understood to provide significant variance in improving student performance. Following that, the presenter will show how several intelligent tutoring systems currently in use are employing mastery learning methods, and how these might be applied to learner feedback in the context of language learning, and in particular, assessment of learning. The presenter will conclude by theorizing about the diminishing role of human tutors in light of these new technologies and consider the implications of this from a historical perspective on technological innovation and adoption in education.