Independent and creative learning in a Digital Electronics course using a web‐based circuit simulator

This paper illustrates the teaching approaches to assessment that foster independent learning in an undergraduate course in the discipline of Electrical and Electronic Engineering. A web‐based circuit simulator of CircuitLab was used to implement some of the laboratory tasks. Practical solutions have been applied to bridge the learning achievement gap between the formative laboratory assessment and the summative examination. To promote creative and independent learning, unique individual self‐discovery projects were given to each student. The projects are generated according to each student's student number in an innovative approach to prevent plagiarism. The scaffolding and self‐discovery lab activities can be modified to be pure online versions, which had been attested in the online course delivery during the COVID‐19 lockdown period in 2020.


| INTRODUCTION
In traditional engineering education practice, laboratory activities are normally well-guided with a detailed laboratory manual [3]. It is true that basic laboratory skills require clear step-by-step instructions. For example, using a digital oscilloscope to observe signal waveforms and measuring the actual peak-to-peak voltages are essential skills and need to be wellexplained in the laboratory manual. However, according to Bloom's learning domains and hierarchy of educational objectives [4], the traditional assessed laboratory tasks target psychomotor skills and only the lower one or two levels of the cognitive domain (knowledge and comprehension in relation to the task). The step-bystep instructions in formative laboratory assessment represent teacher-centred instructivist approaches [6], where the teacher directs and students take a passive role. Although instructivist approaches are suitable to learn some routine or basic skills, they tend to foster student dependency and are poorly suited to achieving in-depth knowledge and understanding. Hence, despite doing well on the formative laboratory assessments, student performance is normally poor in the exams where they need to demonstrate deep understanding and independent application of the learned concepts. Thus, changes are needed to bridge the learning achievement gap between formative laboratory assessments and summative examination.
In contrast, a constructivist approach is able to shift the laboratory tasks from a teacher-centred, passive and dependent approach to a more learner-centred, teacher-as-facilitator, independent learning approach [11]. In the constructivist perspective, learners construct their own knowledge, and the teacher guides and present learning opportunities to facilitate student engagement and learning [6]. Accordingly, laboratory assessments should include some more open-ended tasks, encouraging self-discovery where students could actively experiment, select and apply different strategies, evaluate their outcomes, problem-solve and learn from their mistakes. These tasks would therefore assist the students to become more independent as they would have the opportunity to discover new principles on their own [8].
However, in the context of electrical and electronic engineering education, the major undermining factor in learning and teaching in the laboratory is limited teaching resources, for example, equipment shortages [15]. Traditional laboratory teaching practice involves costly equipment, consumables and maintenance, thereby hindering the educators' intention of promoting independent learning in the laboratory. Thus, researchers have attempted many approaches to promoting independent learning using various technologies. The remote laboratory is a cost-effective way to increase lab accessibility without physical limitations for student learning, such as social distancing, and hardware constraints [9,12]. For example, a semi-virtual electronic laboratory experiment can be implemented by using a webcam to capture the real equipment's real-time signal waveforms controlled by student remote inputs [16]. This remote system is beneficial for students who cannot access the physical laboratory space; however, the system has limited functionality and students cannot change the hardware freely for their own designs. Other attempts include augmented reality using a smartphone [1], and 3D virtual experimentation [15], but they share similar limitations of design freedom.
Web-based software educational tools can be a suitable solution to remotely teach electronic circuits by using some online visual circuit simulators [10]. Customised software tools can also be developed by educators to fulfil their specific needs using software development tools offline, such as Java applets [10], Matlab Simulink [13] and SpiceGen [2]. However, an online circuit simulator tool is preferable for independent learning, as students can try their designed ideas without limitations. Although traditional circuit simulators, such as PSpice and LTSpice are powerful and capable of designing complex electronic circuits, they are hard to use for beginners. This research was carried out using Circui-tLab, which is a web-based electronic circuit simulation programme. The software is user-friendly for students with a limited background in electronic simulation. It allows students to do their circuit design projects anytime and anywhere as long as they have a computer and Internet access. This has proven to be a very effective way for students to practice independence and be in control of their learning pace.
The research questions in this paper are as follows: (1) How can independent learning be incorporated into laboratory tasks using an online circuit simulation tool? (2) Can self-discovery tasks in CircuitLab promote independent learning? (3) How can open-book take-home design tasks in a Digital Electronics course be designed to ensure every student has a unique task of similar difficulty to minimise plagiarism?

| DESIGN AND IMPLEMENTATION
This section describes the innovative process of improving the student learning experience in an undergraduate electronics course-Digital Electronics. It is a core course in the electrical and electronics engineering major and aims to give the students a solid grounding in fundamental electronics and the ability to design electronic circuits for real-life applications. It is taken by approximately 40 second-year engineering students each year. The course comprises a range of assessment tasks, both formative and summative. Formative assessments in the form of laboratory reports serve the dual purposes of providing opportunities for students to build their skills, understanding and learning independence, as well as monitoring how the students' understanding of the course topics is developing. This understanding is later tested summatively via a series of quizzes on the course topics.
Observation of student performance in the course offerings before 2017 indicated that the majority of the students were not independent in their learning. This was reflected in most students achieving good grades in their group-based and well-supported formative laboratory assessment tasks, but very low grades in their summative closed-book tests where they needed depth of understanding and independence to generate solutions without scaffolding. All the formative laboratory assessments in the course had detailed step-by-step instructions, and while they were hands-on tasks, they were not active 'minds-on' tasks.
In consideration of Vygotsky's Zone of proximal development [14], it was recognised that to successfully move from dependence to independence, from closed assessment tasks to open-ended tasks, the students would ZHU AND HOWELL | 635 still require suitable scaffolding tasks to learn basic laboratory skills before moving to more complex tasks [7]. The Christmas Light Controller in the following section is an example of a scaffolding project where students practice basic electronic knowledge discussed in the course and learn relevant hands-on laboratory skills.

| Scaffolding project: Basic Christmas light controller
If the prior knowledge and skills required in the learning process are missing, the learners cannot proceed to more complex tasks. Therefore, to allow students to gain the underpinning knowledge and skills relevant to flip-flops, ring counters and latches, step-by-step tasks were designed and implemented in a scaffolding Christmas light laboratory project. In this scaffolding task, students were asked to implement a 4-bit ring counter, which has been taught in the lectures. Examples of circuit simulation and breadboard implementation results are shown in Figure 1.
These scaffolding tasks are essential for students to learn basic knowledge and laboratory skills. They are hands-on, but passive and dependent learning, which is not enough for students to gain deep learning of the course content. Therefore, the key issue to be addressed is to find an optimal balance point between dependent and independent laboratory tasks. The scaffolding approach with step-by-step instructions should allow students to master the basic skills but be gradually removed as the tasks become more complex. Accordingly, a set of self-discovery design tasks were developed to build on the scaffolding tasks and ensure the students can independently learn more about circuit functions.

| Self-discovery project 1: Bidirectional Christmas light controller
As shown in the lab instruction statement in Figure 2, the students are asked to redesign their circuit to build a bidirectional 5-bit ring counter, which is new to them, and they cannot directly find the solutions in the lecture notes or the course textbook. The new ring counter can automatically control five LEDs in an ON/OFF sequence moving from left to right and back. This is an interesting opportunity for students to link the ring counter learned in the course to the real-life applications of Christmas light controllers.
Open-ended design tasks are essential for deep learning, self-learning and application of the learned knowledge. They are minds-on and promote active and independent learning which often leads to creative solutions. Multiple solutions have been proposed by students, with some of them being very well-designed. Figure 3 shows a circuit diagram and simulated digital waveforms from a student-designed solution.
Some practical limits such as the availability of laboratory space and time must be considered carefully when planning to introduce more self-discovery assessment items into a course. Accordingly, the hardware circuit implementation tasks have been F I G U R E 1 Scaffolding lab task design and results of a Ring Counter in CircuitLab and on Breadboard F I G U R E 2 The laboratory instruction statement for the automatic Christmas light controller design task limited to approximately 60% of the overall lab tasks due to laboratory availability, with the remaining 40% allocated to software-based self-discovery assessment tasks.
The design tasks were given to the students 1 week before the laboratory sessions, so the students would have plenty of time to investigate all possibilities to design their circuits. At the beginning of the corresponding lab session, students needed to show the lab demonstrators their designs in software and explain the functionality of the circuits they had designed. This approach allowed the students to demonstrate their understanding, and the discussion between demonstrator and student helped the students to engage in learning conversations as suggested by socioconstructivist principles [5]. Developing the ability to evaluate and analyse is important to learning independence, and therefore, students are required to compare the results between their simulated circuits and the circuits built in the laboratory and discuss any discrepancies between them. In this way, independent learning can happen despite the limitations of needing time and space in the laboratory.

| Self-discovery project 2: Student Number Counter
One possible issue with self-discovery design assessment tasks is plagiarism. As the design project consists of open-book and take-home tasks, students may copy ideas from each other, defeating the aim of facilitating learning independence. To address this issue, students must provide in-depth explanations and justifications on the how and why of their design choices, and clear marking criteria were provided with detailed weightings for each question. In addition, the lab report is submitted electronically and undergoes a similarity check in Turnitin.
Another more effective way to stop plagiarism is to include unique individual design tasks for each student. However, this may be impractical as many different tasks of similar difficulties would need to be designed, and this would significantly increase the instructor's workload. However, we proposed and implemented an innovative practical solution to generate unique design tasks for each student. For example, in the final lab in this course, students need to apply the scaffolding knowledge and skills gained in the previous labs in more complex design tasks. Students were asked to design and implement a sequential counter using elementary logic gates and flipflops, which can count decimal numbers in a unique counting sequence. The sequence is derived from each student's unique student ID number, following specific rules as follows: (1) Start from 0; (2) From left to right of the student number, keep the numbers that are not repeating; (3) Discard the repeating numbers (including 0) in student number; (4) Add the decimal numbers not shown in student number in the rest of the sequence (from small to large numbers One example of the student number counter is shown in Figure 4. As it is derived from the student's ID number, the counter circuit design and results are unique to the student. This ensures each student has a different sequential digital circuit to work on, and the circuits are at a similar difficulty level, ensuring every student can be assessed fairly. After verifying the designed circuit in the above simulation, the students were requested to explain their circuit functionality to a lab demonstrator and implement their designed circuit in Quartus software and programme it on an field-programmable gate array (FPGA) circuit board in the laboratory. The counting state sequence determined by the student number had to be shown on a seven-segment display on the FPGA board, triggered by a push button. The up and down directions are controlled by a toggle switch on the FPGA board.

| STUDENT FEEDBACK
To evaluate the effectiveness of the independent learning and teaching practice in this course, an online survey was emailed to all students in the course at the end of Trimester 1, 2020. The survey was announced on the course website and an email was sent to all students in the course. The students had up to 2 weeks to input their answers and could alter their answers any time before the due date. Approximately half of the students (19/40) in the class completed it. The survey contained five questions (four quantitative and one qualitative) and students could choose from a range of responses as shown in Table 1.
Questions 1 and 2 were related to the impact and satisfaction with the introduction of self-discovery design tasks in comparison with the traditional step-by-step lab tasks. As shown in Figure 5, the majority of respondents gave positive feedback on the self-discovery projects. 1. The self-discovery design tasks (mini projects, e.g., Christmas light controller, student counter, elevator controller, etc.) were interesting and inspired me to learn valuable skills.
Five items on a Likert Scale.
Strongly Agree to Strongly Disagree.

3.
In comparison with the traditional step-by-step lab tasks, the self-discovery design tasks promoted my independent learning, and engaged me in deep learning in this course.
Five items on a Likert Scale.
Strongly Agree to Strongly Disagree.

5.
Regarding the traditional step-by-step lab tasks and the Selfdiscovery design tasks in all labs.
(a) I wish to have more traditional step-by-step tasks to help me learn basics. 94.4% of the students thought the project was interesting and inspired their learning, and 88.8% thought that independent learning and deep learning were promoted by the self-discovery projects. These results are encouraging and proved the success of the introduction of selfdiscovery projects in this course.
To further investigate the effectiveness of the project task designs, Questions 3 and 4 on the survey were designed to investigate how students felt about the quantity and difficulty of the self-discovery design task. The feedback results shown in Figure 6 were very positive, with 83.3% of the students agreeing that the step-by-step lab tasks and the self-discovery design tasks were well balanced, and 66.7% agreed that the design task difficulty was appropriate and fair.
Question 5 was an open question for student anonymous comments. A total of 14 detailed comments were received, and most of them are positive and constructive. The students appreciated the wellbalanced basics and deep learning lab tasks, with sample comments such as the fundamentals were taught well and it then gives you a chance to apply your knowledge.
Some students felt some difficulty in doing the selfdiscovery tasks and were frustrated at the time, however it also was very rewarding when I understood what I was missing and the simulations worked. Some students also found the CircuitLab tasks built confidence, commenting that The self-discovery tasks were great. They allowed me to prove to myself I could solve complex problems with the knowledge I gained from this course. This has given me the confidence in myself that I can be a successful electrical engineer in industry in the future. The effectiveness of the independent learning and teaching approaches since the implementation in 2017 can be evidenced by other channels, for example, student achievements and Student Experience of Course (SEC) scores. After the implementation of the research method in 2017, the student's median overall mark improved from 62% (2016) to 70% (2017) and kept consistently high at around 70% in 2018 (73%), 2019 (69%), 2020 (74%) and 2021 (71%). The average enrolment was 47 during the period from 2016 to 2021. Meanwhile, the marks gap between lab work and closed book tests decreased from 20.2% (2016) to 15.0% (2017). In addition, the SEC score (Question 6) has increased significantly from 3.5 (2016) to 4.5 (2017), and maintained steady improvement to 4.7 (2018) and 4.9 (2019) as more independent learning activities were introduced over the years, as illustrated in Figure 7.

| CONCLUSION AND DISCUSSION
The successful results since 2017 evidenced that independent learning and teaching approaches can bridge the student learning achievement gap between formative laboratory assessments and summative examinations. Both scaffolding and self-discovery lab assessments are important and valuable in learning, and it is, therefore, essential to balance them in a proper way, with suitable levels of difficulty and appropriate assessment weighting within the course. Students believed the self-discovery tasks promoted independent learning and deep learning, although they indicated it can be challenging in the beginning. More self-discovery and unique tasks of various levels of difficulty will be beneficial in future offerings of the course. An example task could be to design an elevator controller for a five-story building, which can control the up or down direction and count the floor numbers. However, it should be noted that more unique design tasks will increase the marking load as the markers have to carefully investigate and check each student's unique design solution.
The above-mentioned scaffolding and self-discovery lab activities can be modified to be pure online versions, which had been attested in the online course delivery during the COVID-19 lockdown period in 2020. In response to the impact of the COVID-19 pandemic in 2020, we shifted the physical lab activities to online alternatives which can be fully implemented in the online circuit simulator of CircuitLab. To achieve the same learning outcomes, the contents in lab tasks were kept no change, but we included detailed and visualised instructions on how to set up the lab experiment in the online environment. In addition, we accommodated more online design activities in the online course delivery. For example, some additional counting functions of counting upwards and downwards had been introduced in the Student Number Counter project in 2020; some more design tasks, for example, a mini project for an elevator controller were added. Thanks to the experiences gained since 2017, the COVID-19 online delivery ran smoothly, and student feedback and achievement remained as good as the pre-COVID course deliveries: in 2020, the median overall mark was kept consistently high at 74%, and the SEC score was also kept consistently high at 4.5 (out of 5).

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.