EFFECTIVENESS OF VISUAL PROGRAMMING AND TEXT-BASED PROGRAMMING IN DEVELOPING STUDENTS' COMPUTATIONAL THINKING SKILLS

Abstract

Computational thinking is an essential skill in 21st-century education because it supports logical, systematic, and structured problem solving. This study compared visual programming using Scratch and text-based programming using Python in developing students' computational thinking skills. A quantitative quasi-experimental pretest-posttest comparison-group design was conducted at SMK N 4 Kupang with 25 students in Class X DPK and 24 students in Class X DKV 1. Each group received one instructional session scheduled for 2 × 45 minutes, with the same researcher serving as instructor. Both groups learned standard algorithms and the Stack data structure through equivalent learning objectives, materials, activities, tasks, instructional sequence, duration, and assessment procedures. Data were collected using an essay-format computational-thinking test and classroom observations. Descriptive statistics and N-Gain scores were used to describe learning improvement, while an independent-samples t-test was used to compare N-Gain scores. The visual programming group increased from a mean pretest score of 45.40 to 88.13, with an average N-Gain of 0.78 (High), whereas the text-based group increased from 48.00 to 72.50, with an average N-Gain of 0.49 (Medium). The between-group comparison yielded t(47) = 8.12, p < .001, with Cohen's d = 2.37. Classroom observations indicated more structured use of visual blocks for algorithm construction in the Scratch group and more syntax-focused trial-and-error behavior in the Python group. The findings are interpreted within the context of two intact classes at one school.