At an unrivaled and enduring pace, computing has transformed the world, resulting in demand for a universal fourth foundation beyond reading, writing, and arithmetic: computational thinking (CT). Despite increasingly widespread acceptance of CT as a crucial competency for all, transforming education systems accordingly has proven complex. The principal hypothesis of this thesis is that we can improve the efficiency and efficacy of teaching and learning CT by building gameful learning and assessment systems on top of block-based programming environments. Additionally, we believe this can be accomplished at scale and cost conducive to accelerating CT dissemination for all.
After introducing the requirements, approach, and architecture, we present a solution named Gameful Direct Instruction. This involves embedding Parsons Programming Puzzles (PPPs) in Scratch, which is a block-based programming environment currently used prevalently in grades 6-8. PPPs encourage students to practice CT by assembling into correct order sets of mixed-up blocks that comprise samples of well-written code which focus on individual concepts. The structure provided by PPPs enable instructors to design games that steer learner attention toward targeted learning goals through puzzle-solving play. Learners receive continuous automated feedback as they attempt to arrange programming constructs in correct order, leading to more efficient comprehension of core CT concepts than they might otherwise attain through less structured Scratch assignments. We measure this efficiency first via a pilot study conducted after the initial integration of PPPs with Scratch, and second after the addition of scaffolding enhancements in a study involving a larger adult general population.
We complement Gameful Direct Instruction with a solution named Gameful Constructionism. This involves integrating with Scratch implicit assessment functionality that facilitates constructionist video game (CVG) design and play. CVGs enable learner to explore CT using construction tools sufficiently expressive for personally meaningful gameplay. Instructors are enabled to guide learning by defining game objectives useful for implicit assessment, while affording learners the opportunity to take ownership of the experience and progress through the sequence of interest and motivation toward sustained engagement. When strategically arranged within a learning progression after PPP gameplay produces evidence of efficient comprehension, CVGs amplify the impact of direct instruction by providing the sculpted context in which learners can apply CT concepts more freely, thereby broadening and deepening understanding, and improving learning efficacy. We measure this efficacy in a study of the general adult population.
Since these approaches leverage low fidelity yet motivating gameful techniques, they facilitate the development of learning content at scale and cost supportive of widespread CT uptake. We conclude this thesis with a glance at future work that anticipates further progress in scalability via a solution named Gameful Intelligent Tutoring. This involves augmenting Scratch with Intelligent Tutoring System (ITS) functionality that offers across-activity next-game recommendations, and within-activity just-in-time and on-demand hints. Since these data-driven methods operate without requiring knowledge engineering for each game designed, the instructor can evolve her role from one focused on knowledge transfer to one centered on supporting learning through the design of educational experiences, and we can accelerate the dissemination of CT at scale and reasonable cost while also advancing toward continuously differentiated instruction for each learner.
| Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/3n01-n276 |
| Date | January 2023 |
| Creators | Bender, Jeff |
| Source Sets | Columbia University |
| Language | English |
| Detected Language | English |
| Type | Theses |
Page generated in 0.0023 seconds