Attending to structural programming features predicts differences in learning and motivation

• Published in: Journal of computer assisted learning 2018-04, Vol.34 (2), p.115-128
• Main Authors: Witherspoon, Eben B., Schunn, Christian D., Higashi, Ross M., Shoop, Robin
• Format: Article
• Language: English
• Subjects: CAI; Classrooms; computational thinking; Computer assisted instruction; Computer science; Computer Science Education; Curricula; Direct Instruction; Education; Educational Environment; Educational psychology; Educational technology; Instructional Effectiveness; Knowledge; Knowledge management; Learning; Logic; Microprocessors; Middle School Students; Motivation; Programming; Robotics; School environment; Skills; Student Motivation; Transfer of Training; Units of Study
• ISSN: 0266-4909; 1365-2729
• DOI: 10.1111/jcal.12219

Educational robotics programs offer an engaging opportunity to potentially teach core computer science concepts and practices in K–12 classrooms. Here, we test the effects of units with different programming content within a virtual robotics context on both learning gains and motivational changes in middle school (6th–8th grade) robotics classrooms. Significant learning gains were found overall, particularly for groups introduced to content involving program flow, the structural logic of program execution. Relative gains for these groups were particularly high on items that require the transfer of knowledge to dissimilar contexts. Reaching units that included program flow content was also associated with greater maintenance of programming interest when compared with other units. Therefore, our results suggest that explicit instruction in the structural logic of programming may develop deeper transferrable programming knowledge and prevent declines in some motivational factors. Lay Description What is already known about the topic: Educational robotics may provide an engaging learning environment for computer science (CS); Teaching generalizable programming and CS to all students is an emerging area in K–12 curricula; and Limited research exists relating features of robotics learning environments to motivational factors and the development of generalizable programming knowledge. What this paper adds: Virtual robotics provides a rich learning environment for developing programming knowledge and skills; Programming tasks that shift superficial features but attend to similar structural features of the program provide larger gains in generalizable programming for learners; and Participation in more conceptually rich programming tasks maintains learner interest but has differential effects on learners' ability beliefs. Implications of the study findings for practitioners: Virtual robotics curriculum can offer a productive learning context for K–12 CS courses that aim to teach generalizable programming knowledge and skills; Explicit attention to the structural logic of programming may develop programming knowledge and skills that can be generalized beyond the immediate learning context; and Providing all learners with opportunities to engage in well supported but conceptually challenging programming units may prevent declining interest in middle school CS.