Student Autonomy Improves Learning: A Theoretical Justification of the Empirical Results

In many pedagogical situations, it is advantageous to give students some autonomy: for example, instead of assigning the same homework problem to all the students, to give students a choice between several similar problems, so that each student can choose a problem whose context best fits his or her experiences. A recent experimental study shows that there is a 45% correlation between degree of autonomy and student success. In this paper, we provide a theoretical explanation for this correlation value. 1 Formulation of the Problem Traditional non-autonomous approach to learning. Traditionally, all the students in the class are assigned the same homework problems, the same practice problems, and are given the same problems on the tests. Need for autonomy. The main advantage of the traditional approach seems to be that all the students are treated equally: they are given exactly the same problems, so they have the same chance to succeed. In reality, however, the traditional approach has its limitations. For example, in a university setting, every engineering student has to take calculus. Engineering students usually form a majority in calculus classes. However, in addition to engineering students, also students from other disciplines, e.g., from bioinformatics, are required to take calculus. Since the majority of students in a calculus class are engineering students, most textbook application-related problems come from engineering or from related fields. This fact gives an unfair disadvantage to bioinformatics students many of whom are not very familiar with the main concepts from engineering. As a result, it is much harder for a bioinformatics students to study in this class, and for those bioinformatics students who worked hard and mastered all the concepts, it is more difficult to show their knowledge on the tests – since