Does text-based programming improve K-12 students’ CT skills? Evidence from a meta-analysis and synthesis of qualitative data in educational contexts

Promoting computational thinking (CT) skills in programming activities has become a key issue in K-12 education, and the shift from visual to text-based programming is an inevitable trend in programming education. Although research on K-12 text-based programming to promote CT skills has been increasing, a literature review that systematically organizes and analyzes it has not yet emerged. Moreover, there are no uniform conclusions on how to design text-based programming activities to effectively promote students' acquisition of CT skills. The purpose of this study was mainly to determine the effectiveness of using text-based programming to enhance K-12 students' CT skills, to explore the effects of various text-based programming instructional design factors on CT skill acquisition, and to discuss the key issues and future development of text programming in the K-12 stage. In this study using the quantitative method of meta-analysis and the qualitative method of systematic literature review, 46 studies of text-based programming on K-12 students' CT skills were retrieved, of which 19 empirical studies including 37 effects size met the inclusion criteria for the meta-analysis and were therefore used as a meta-analysis. The results showed that text-based programming improved the CT skills of K-12 students overall (Hedges' g = 0.71, 95% CI [0.51, 0.90], p < 0.01). No publication bias was detected. In addition, the results of the moderator analysis revealed that: 1) text programming had the greatest impact on computational thinking development in middle school students (g = 0.833, p = <0.05); 2) the duration of text programming instruction had the most significant positive impact on students' CT at 16–20 weeks (g = 1.038, p <0.05); 3) the moderating effect of different assessment tools on CT skills in text programming instruction The difference was statistically significant, and the test assessment method had a highly significant positive moderating effect on CT skills of students in grades K-12 (g = 1.005, p < 0.05); 4) The moderating effect of different student sample sizes on students' computational thinking development was significantly different, and the implementation of text-based programming instruction had the most significant effect on students' CT improvement among students with sample sizes of 41 ∼ 60 (g = 0.921, p < 0.05); 5) There were no significant differences in the effects of different text-based programming languages and learning environments on students' computational thinking.However, there are still barriers to developing instructional practices for text-based programming to foster CT in K-12 students, such as the highly abstract and complex syntax of text-based programming. This study found that the most critical issues that need to be addressed in current text-based programming to foster CT in K-12 students are the lack of student interest in text-based programming and the misconceptions that arise when students move from visual programming to text-based programming. To address these issues, future research could begin with improving teaching methods and increasing student motivation and interest. To address these issues, future research could begin with improving teaching methods and increasing students' motivation and interest.