Learning Computational Thinking Efficiently with Block-based Parsons Puzzles

Abstract

To investigate learning system elements and progressions that affect computational thinking (CT) learning in block-based environments, we developed a Parsons Programming Puzzle (PPP) module within Scratch with scaffolding customized via a novel Blockly grammar. By varying the presentation and types of feedback encountered between- and within-subjects in a study of 579 adults, we identified features and scaffolding strategies that yield manageable cognitive load (CL), improved CT learning efficiency, and increased motivation, for a general populace. Findings indicate: 1) PPPs with feedback induce lowest CL; 2) an isolated palette, correctness feedback, and fading correctness feedback increase learning efficiency; 3) fading scaffolding can increase CT motivation. We analyze 12 conditions to provide insight to those developing block-based PPP systems with the aim to advance equitable CT education for all.