In this Work in Progress paper in the Research Category we report on existing concerns about Scratch programming, and introduce patterns as a possible solution. Scratch is a popular language for introducing students to programming, but there is a concern that the students might not be exposed to all the key elements of programming when the development environment tempts them to explore elements such as the range of sprites available. We propose the use of programming patterns as a measure of the sophistication of student work. To understand the importance of patterns we report on our initial work that analyzes a large number of projects from the public Scratch repository to evaluate how extensively the basic patterns appear in student work. This can help inform the improvement of teaching methods to include use of broader range of patterns.

This Full Paper in the Research Category evaluates the use of remixing in Scratch. A feature of the Scratch programming environment is that it supports students to share their code and "remix" (modify) other students' code. Remixing in Scratch has garnered much attention by the research community as use of collaboration for learning was one of the main ideas behind Scratch. It can provide opportunities to read others' code, learn how features can be implemented using the Scratch language, and contribute to the program. It can also prevent students from engaging with the code if they copy an existing program that does what they are trying to do without needing modification. The literature shows mixed results regarding use of remixes in Scratch. We have investigated at a large scale what happens in practice by analysing thousands of student programs shared through the Scratch online repository. As well as replicating prior work on a larger scale to show the impact of remixing on learning programming skills through Lines of Code (LOC) and repertoire of block usage, we also measure the use of elementary patterns (common combinations of commands). We track the progress of each project through its remixes and compare the results between the root version and the final version.

No abstract

Block based programming languages are a popular way to introduce computer programming to young students. However, concerns have been raised that students may not be learning important programming skills. Previous analyses of Scratch projects online have revealed that the use of important computational elements such as variables, if statements and controlled loops is extremely low, and introducing elementary patterns to the teaching sequence emerged as a potential solution. Although students might have success in the Scratch environment with a limited vocabulary of constructs, it has been pointed out in the literature that the transition process to a different programming language might be difficult. A concern with prior studies of a static analysis of uploaded Scratch projects is that they might be biased towards one-off or brief encounters with programming, where students have a short experience, upload a simple program, and share no further work. We address this concern in this study by measuring the progress of Scratch users over the years, particularly through their use of elementary patterns. Our study has identified all projects submitted by 34,832 users in the Scratch community, and for each user we have compared the skills demonstrated in the first half with those in the second half of their projects. There were no clear signs of any progression, and even some indications of negative progression, and even after more than a year of programming, for most users we found that the use of key elements of programming was still low.

No abstract