- The paper introduces the CTSkills app, a web-based tool designed to assess K-12 students' problem decomposition skills, a critical yet often overlooked aspect of computational thinking.
- The study used the CTSkills app with 75 students across grades 4-9, finding significant variations and improvements in substantive and relational decomposition skills as students advanced through grades.
- The CTSkills app offers a scalable method for assessing decomposition, with findings suggesting a need for balanced CT instruction and recommending future longitudinal studies.
Assessing Problem Decomposition Skills in Computational Thinking: Insights from the CTSkills App
The paper "The CTSkills App - Measuring Problem Decomposition Skills of Students in Computational Thinking" presents a study addressing a key challenge in K-12 computer science education—assessing students’ problem decomposition skills, a foundational aspect of computational thinking (CT). The study introduces the CTSkills app, a web-based assessment tool designed to provide insights into how students from primary to secondary education break down complex problems into manageable sub-problems.
Core Findings and Methodology
The authors argue that problem decomposition is often underemphasized in favor of algorithmic solution formulation within educational contexts. Responding to this gap, the CTSkills app was developed to measure decomposition proficiency, gathering data from 75 students across grades 4-9. This application features interactive game-based scenarios where students are tasked with identifying relevant objects, assigning properties, and establishing relationships between different virtual elements. The assessment is heavily grounded in a framework proposed by P.J. Rich et al., which categorizes decomposition into substantive, relational, and functional types.
The research demonstrates a significant variation in decomposition skills across different education levels. Notably, the study found that the ability to accurately identify relevant targets and avoid non-target elements improved with the students' grades. The findings underscore a marked improvement in substantive and relational decomposition skills as students age, although discrepancies within specific grade levels were observed.
Statistical Analysis and Results
The study employs robust statistical methods to validate its findings, including ANOVA, Chi-Square Tests of Independence, and Linear Mixed-Effect Models (LMMs). The lack of significant gender differences in scores indicates equitable skill development opportunities across genders in CS education contexts. Furthermore, the study identifies specific grades, such as 8, where decomposition skills peak, whereas a decline is observed in grade 9, suggesting possible transitional academic challenges at this stage.
Implications for Computational Thinking Education
The research presented has several significant implications for educators and curriculum developers. The CTSkills app provides a scalable and automated method for assessing decomposition skills in a classroom setting, highlighting areas where instructional focus could be improved. As computational thinking becomes increasingly intertwined with K-12 curricula, understanding baseline decomposition abilities can inform teaching strategies that emphasize fundamental CT skills from an early age.
Future Directions
The study lays the groundwork for continuous refinement of the CTSkills app, suggesting expansions in task complexity and scope, particularly toward functional decomposition tasks suitable for older students. Future research is recommended to employ longitudinal study designs to trace the development of decomposition skills over time, offering insights into the cognitive aspects of CT in educational settings.
Ultimately, the paper advocates for refined pedagogical approaches that balance the instruction of core CT components, fostering comprehensive problem-solving skills among students equipped to navigate and shape an increasingly digital world.
