Kids teaching Kids: Peer tutoring of programming languages

DE LA HERA, DP; SIGMAN, M; CALERO, CI

Amsterdam

Congreso; 59th Annual Meeting Psychonomic Society; 2018

Psychonomic Society

Teaching is regarded as a natural cognitive ability which we are all exposed to and engaged in. Surprisingly, it has been much less studied by neuroscience and cognitive science than learning, its counterpart. Particularly, teaching among companions of equal status, or peer tutoring (Topping, 2015), which may play an important role in children´s cognitive development, has been even less studied by experimental psychology or neuroscience. Therefore, there is little to no quantitative evidence about how children engage in peer tutoring and about its impact on both tutees and tutors.We propose here that a first step into understanding how peer tutoring may help children, both tutor and tutees, understand and learn new skill. In particular, we studied programming languages and their fundamentals, tackling some of the challenges implied in teaching these undoubtedly important skills in today´s world.In the present study we aimed to compare i) learning in a regular class set-up, with one adult teacher and a group of children, and ii) peer tutoring, with one children teaching another of around the same age. We targeted differences in performance, as well as in strategies used to solve the test problems presented.In addition, we hypothesize that revision and reorganization of knowledge, supposedly involved in teaching, will have an impact on tutor performance, setting them aside from their classmates who did not teach. Our approach was designed accordingly to quantify this impact.The study consisted of four stages. First, in the Instruction stage, groups of second-grade children were taught by an adult how to use a LEGO programming language designed for children. Second, in the Test stage, we measured these children´s proficiency. Third, in the Interaction stage, some of those children (tutors) had to teach how to use the new language to inexperienced classmates (students). A group of children who had participated in the first stage did not interact and instead revised the pedagogical material on their own, acting as control group. Finally, in the Retest stage, in which all children were tested.To our knowledge, this is the first systematic and quantifiable study pertaining peer tutoring and learning of programming skills, both priority research areas in the field of education. Analysis of our results have shown that children perform equally well, irrespective of whether they learned from an adult or from a peer. Further analysis may inform about more subtle differences between both conditions, and also about possible tutor benefits not present in children who did not go through the hypothetically enriching experience of teaching.Further analysis of our data will also allow us to identify strategies and resources children use to teach, which we propose may be related with both tutor and tutee performance.