Abstract:

Quantum computing is an emerging paradigm with increasing relevance for optimization and machine learning, yet its mathematical formalism often becomes a barrier for learners in advanced computing programmes. As a result, students may reproduce circuit recipes without developing a robust conceptual model that links abstract principles to observable behaviour. Building on previous work that introduced guided, hands-on quantum programming activities, this paper presents an instructional approach that strengthens conceptual understanding by combining simulation-based learning with executions on cloud-accessible quantum hardware.

The learning design is structured around an active “predict–run–explain” sequence supported by Python-based quantum programming with Qiskit. Students first articulate qualitative predictions about circuit outcomes (e.g., expected measurement distributions and the effect of specific gates), then implement and run the circuits, and finally explain results by connecting observations to key concepts such as superposition, measurement, entanglement, and interference. The approach is intentionally scaffolded: tasks progress from single-qubit intuition-building exercises to multi-qubit circuits and simple algorithmic patterns, with formative checkpoints that prompt students to justify their reasoning and refine misconceptions.

Rather than treating non-ideal results as errors to ignore, students are asked to compare idealized simulation outputs with real-device executions and to interpret discrepancies through concepts such as noise, sampling variability, and hardware constraints. This comparison is embedded in the lab workflow and assessment criteria, encouraging students to develop realistic expectations about near-term devices and to adopt a more scientific stance toward experimentation, uncertainty, and model limitations.

The implementation was carried out in a postgraduate Artificial Intelligence (AI) master’s context and evaluated through evidence from practical artefacts (lab notebooks and programming tasks), short conceptual checks, and student feedback on perceived understanding, engagement, and relevance. The results show clear improvements in students’ ability to connect formal concepts with observable circuit behaviour, increased confidence when working with quantum programming tools, and higher perceived authenticity of learning when real hardware is introduced. These findings reinforce the value of combining simulation and real-hardware experimentation as a practical pathway to introduce quantum computing in postgraduate curricula.

Keywords:

Quantum computing education, Active learning, Simulation-based learning, Qiskit, Artificial intelligence education.