AlgoQuest: A Web-Based and Mobile-Based Educational Game for Teaching Algorithmic Thinking

Abstract: Game-based learning can close the gap between what students hear in lectures and what they can actually do when they write algorithms. In practice, though, educational software runs into three recurring problems: keeping progress in sync across devices, getting acceptable performance on phones as well as desktops, and keeping learners with very different backgrounds motivated. This paper describes the design, implementation and evaluation of AlgoQuest, an educational game that runs in web browsers (HTML5/WebGL) and on Android and iOS from one shared codebase. A lightweight serverless backend stores progress. A Dynamic Difficulty Adjustment (DDA) engine reads each learner's completion time, then adjusts puzzle difficulty and the amount of scaffolding. We tested the game in a four- week controlled study with 80 MCA students, using a pre-test/post-test design. The experimental group improved more than the control group (Hake's normalized gain g = 0.581 vs. 0.297; p < 0.0001; Cohen's d = 1.81). The experimental group also reported much lower frustration on the NASA-TLX and rated usability at 81.42 on the System Usability Scale. On a mid-range Android phone the game held about 58 FPS, with cloud synchronization under 180 ms. These results suggest that a single architecture can support both browser and mobile delivery of an algorithm-learning game. Because the study covers one department and one term, the results should be read as promising rather than final.

Keywords: digital game-based learning, cross-platform architecture, WebGL, dynamic difficulty adjustment, flow theory, System Usability Scale, instructional technology.