1. Introduction
With the global rise of the maker movement, governments across the world have begun to focus on the impact of maker activities on learners. The philosophy of maker education is learning by making; this indicates a transformation from the conventional dissemination of knowledge to learning relevant concepts by doing. That is, maker education allows students to implicitly acquire knowledge while completing maker projects. This method of learning has a positive and effective impact on learners. Compared with conventional teaching methods, students think about and apply their knowledge, during which they proactively identify and address problems instead of acquiring knowledge passively.
Scientific, manufacturing, and other technological advances have reduced the costs of maker equipment. Also, as the maker movement has developed, embedded development systems have matured, and sensors have become more affordable and diverse, helping learners develop their creativity. Information technology (IT) has resulted in new development and applications for maker education, and it is now cheaper and easier to integrate maker education into on-site education than it once was. These developments all significantly benefit learners. Countries all over the world have developed maker education, and the U.S. Department of Education is cooperating with Exploratorium in offering maker courses to high-poverty and low-performing regions [
1]. In Europe, the Fabrication Laboratory (FabLabs), the EU-initiated MakerSpace, and organizations such as Maker Faire Rome and Startup Europe provide maker spaces for learners to tap into their creativity. In Taiwan, the movement has been nurtured by the Workforce Development Agency, which has established maker bases and factories. In addition to hardware developments, the maker concept has been brought into the classroom. Maker education requires not only hardware investments but also integrated software, as well as teachers and teaching materials, helping learners to gain a better understanding of the potential value of the maker movement. Furthermore, students can be trained to acquire existing knowledge from tasks.
The literature shows that maker education integrates well with emerging science and technology, such as STEAM education [
2,
3], virtual reality (VR) [
4], and computational thinking (CT) [
5]. This gives learners access to emerging science and technology in a better and faster manner, reflecting today’s rapidly changing culture. Additionally, the literature indicates that maker education assists students in developing their creativity, collaborative skills, and problem-solving abilities [
6] and improves their engagement in classes [
7], all of which have long-term, vertical impacts on learners rather than short-term impacts [
8].
CT has been a popular research topic in recent years [
9]. In 2006, Professor Wing of CMU indicated that CT is a basic skill needed in daily life and that CT is a key element for elementary education. She re-defined CT and showed that it is just as important as the “3 Rs” (reading, writing, and arithmetic) and that every child should be encouraged to hone their analytical skills using CT [
10]. CT is a thinking process in which people use basic concepts and logical methods from computer science to identify and seek solutions step by step [
10,
11]. Accordingly, learning CT helps us tackle problems more effectively, understand root causes, and address more sophisticated problems [
12,
13]. In addition, the increasing importance of CT has motivated countries throughout the world to implement CT training policies [
14,
15].
CT is generally learned through programming [
16,
17]. Though current programming languages closely resemble natural languages, abstract concepts that are implicit in text-based programming languages are difficult for beginners to learn [
18]. In contrast to such text-based programming languages, visual programming languages (VPLs) present language structures via visual blocks of different colors and shapes. This enables beginners to design programs by manipulating blocks, thus significantly lowering the threshold of programming [
19]. Relevant studies demonstrate that VPL is an effective learning method for CT, which explains the increasing use of VPL in CT education using systems such as Scratch [
20] and Blocky [
21]. Scratch facilitates user-defined block-based design to design programs using VPLs and also connects with IoT devices. Therefore, Scratch is the most popular learning instrument for CT [
22]. Despite the fact that many studies have attested to the effectiveness of VPL for learning CT [
22,
23], most determined it by quantitative methods [
24,
25] involving CT tests or scales [
25,
26,
27]. Some assessed student programming projects through operating systems [
25,
28]. However, such methods failed to comprehensively analyze the programming and learning processes and thus did not investigate students’ operations during visual programming (VP). It is also important to effectively and automatically assess the learning effectiveness of CT [
29] and determine whether students understand CT, particularly in problem-solving.
We thus tapped the Scratch VPL as a programming language tool to teach CT and developed a learning and tracking system for CT education. This system facilitated real-time tracking of programming projects and tasks, allowing teachers to grasp the learning pace of every student as well as the various project results. The system logged the writing procedures and paths of students during programming assignments to help teachers diagnose students’ learning weaknesses. Timely intervention and assistance then alleviated students’ anxiety and boosted learning motivation and confidence. We posed the following research questions:
- (1)
After participating in the course of Scratch programming, is there any difference in the frequency of using computational thinking skills?
- (2)
Does participation in the Scratch programming course affect learning motivation, learning anxiety, and learning confidence?
4. Discussion
We developed a CT learning platform and established a learning and tracking system for Scratch. We used the behavior records to better understand the learners’ VPL programming approaches for various tasks. The analysis results helped teachers understand the students’ programming progress and offer timely support when necessary. CT problem-solving not only improved the logical skills and systemic thinking of learners but also enhanced their problem-solving skills, which were equally valuable for problems encountered in daily life.
The FCT analysis showed that most learners understood the four core CT steps and solved the problems based on CT concepts. Nonetheless, when they attempted to solve the problems for the first time, they were not able to master the CT steps. Different attempts to solve the problems and tasks only increased their learning anxiety. Since the learners’ progress was tracked and recorded in the backend database, the teacher was able to know what difficulties each student faced by consulting their logs, and was thus able to render timely assistance and offer helpful instructions. Such assistance relieved anxiety generated during tasks, improving the students’ motivation and confidence.
In addition, learners mastered the four core CT steps and concepts as they completed the tasks. Learning anxiety did occur during the process, but a certain amount of anxiety is necessary to enhance learning efficiency [
35]. Moreover, using the concept of computational thinking could improve the logic and systematisms of learners; learners could also apply computational thinking in the process of solving problems in the future and even effectively achieve the effect of learning transfer. Solving the various problems familiarized students with the CT steps, boosted their confidence, and taught them to keep trying in the face of difficulty. Timely assistance and explanations from the teacher also helped students maintain their motivation and gave them the confidence to complete the tasks.
The results of this study are consistent with the drive theory [
39], in which the intrinsic drive motivates intrinsic physiological needs, resulting in behavior. Motivation is indispensable for learning. When the needs of individuals are not met, the internal drive is stimulated, leading to reactions. Needs must be satisfied to achieve the desired result. Accordingly, teachers consulted the system’s learning logs and analyses to understand the programming progress of each student and offered instruction and assistance to meet their needs. Repeated success in solving problems enhanced learners’ confidence and motivation, and the desire to complete more tasks stimulated internal drive, motivating students to accomplish the tasks. The inverted U-shape theory [
40] also supports these results: there is a U-shape curvilinear relationship between learning performance and anxiety. As anxiety rises, performance will gradually improve, and when anxiety rises to a certain level, the best performance will be produced. In the tasks in this experiment, learner anxiety was maintained at an acceptable range, promoting learners’ confidence and interest in the system activities and helping them to complete the tasks.
5. Conclusions and Recommendations for Future Work
In this study, we developed a CT learning platform and established a learning and tracking system for Scratch. The system tracked and recorded programming activities during the course so that teachers could better understand each student’s progress and difficulties. Using this learning and analysis system, teachers could offer timely assistance and relevant explanations to students at different levels. The analysis results show that all participants were able to complete the tasks and solve the problems according to the four core CT steps. When the students faced difficulties, the teacher offered instructions and assistance to reduce their anxiety. Additionally, various tasks served to familiarize the students with CT, and problem-solving boosted their confidence and motivation.
From interviews, we learned that the course was too fast-paced. As a result, though participants had basic computer and programming skills, they still needed time to familiarize themselves with the four core CT steps for problem-solving. In addition, we learned that the proposed CT learning platform has many potential applications. Specifically, the learning and tracking system is well-suited for higher education. The CT-based problem-solving, pattern recognition, abstraction, and algorithm design can be incorporated into course designs to train students’ logical thinking and problem-solving capabilities.