[Paper Review] Introduction to Computational Thinking: a new high school curriculum using CodeWorld

Fernando Alegre, John Underwoood+2 more|arXiv (Cornell University)|Nov 30, 2019

Teaching and Learning Programming4 citations

TL;DR

This paper presents a high school computational thinking curriculum, Introduction to Computational Thinking (ICT), using the CodeWorld environment to teach programming and mathematical reasoning through a Haskell-based language. The course, designed for ninth graders, emphasizes conceptual foundations over syntax, resulting in a 24% average gain in content knowledge and positive student attitudes despite being perceived as challenging.

ABSTRACT

The Louisiana Department of Education partnered with the Gordon A. Cain Center at LSU to pilot a Computing High School Graduation Pathway. The first course in the pathway, Introduction to Computational Thinking (ICT), is designed to teach programming and reinforce mathematical practice skills of nine-grade students, with an emphasis on promoting higher order thinking. In 2017-18, about 200 students and five teachers participated in the pilot, in 2018-2019 the participation increased to 400 students, and in the current 2019-2020 year about 800 students in 11 schools are involved. Professional development starts with a five-week intensive summer institute, which is complemented with follow-up Saturday sessions and coaching support during the academic year. After describing the course content and briefly the teacher training, we discuss the data we have collected in the last two years. The overall student reception of the course has been positive, but the course was categorized by most students as hard. However, the Computing Attitude Survey analysis indicates that the difficulty of the course did not demotivate the students. The pre-post test content assessments show that students learned not only the language, but also general principles of programming, logic and modeling, as well as use of variables, expressions and functions. Lessons learned during the pilot phase motivated changes, such as emphasizing during PD the need to provide timely feedback to students, provide detailed rubrics for the projects and reorganize the lessons to increase the initial engagement with the material. After two years of running pilots, the course is becoming student-centered, where most of the code and image samples provided in the lessons are based on code created by previous students.

Motivation & Objective

• To develop a computational thinking curriculum that connects programming to algebra, geometry, and science modeling for general high school students.
• To create a teacher training model that enables non-CS teachers (e.g., math and science educators) to effectively deliver the course.
• To design a student-centered, programming-focused curriculum that fosters higher-order thinking and conceptual understanding in underserved populations.
• To evaluate student learning gains and attitude changes in a large-scale pilot across multiple school districts.

Proposed method

• The course uses CodeWorld, a web-based IDE based on a simplified variant of Haskell, to teach programming concepts through visual and mathematical modeling.
• Lessons are structured around Computational Thinking principles, introducing syntax only when needed, with initial focus on expressions, variables, and functions.
• Professional development includes a five-week summer institute where teachers learn by completing student assignments, practicing peer feedback, and designing lessons.
• The curriculum emphasizes abstraction, decomposition, and iterative development through pair programming and collaborative scene creation.
• Assessments include the Conceptual Foundations of Coding Test (pre- and post-test) and the Computing Attitude Survey (CAS) to measure learning and attitude changes.
• Data analysis uses Wilcoxon signed-rank tests to compare pre- and post-assessments and correlation analysis to link attitude and learning gains.

Experimental results

Research questions

• RQ1 To what extent does the ICT curriculum improve students’ understanding of programming, logic, and modeling concepts?
• RQ2 How do student attitudes toward computing change after completing the course, especially given its perceived difficulty?
• RQ3 What impact does teacher professional development have on instructional quality and student learning outcomes?
• RQ4 How do learning gains in computational thinking correlate with student attitudes toward computing?

Key findings

• Students showed a statistically significant average gain of 24% on the Conceptual Foundations of Coding Test, with post-test scores rising from 29.5% to 53.8%.
• The Wilcoxon signed-rank test confirmed the pre- and post-test distributions were significantly different (p < 2.2 × 10^(-16)).
• All four test categories—variables, expressions and functions, CodeWorld specifics, and mathematical modeling—showed higher average post-test scores than pre-test scores.
• The Computing Attitude Survey showed minimal change in student attitudes (shift of 0.02), with strong pre- and post-test correlation (R = 0.54, p = 1.3 × 10^(-10)).
• A positive correlation (R = 0.29, p = 0.001) was found between students’ post-course attitudes and their learning gains on the content test.
• Despite being uniformly rated as hard by students, the course did not demotivate learners, and the majority reported positive experiences.