The study, published in the Journal of New Approaches in Educational Research, was motivated by a sobering national context. Colombia has one of the highest university dropout rates in Latin America, and engineering programs have been particularly hard hit, with some institutions closing programs due to low demand and the country facing a shortage of qualified engineers. Researchers have documented widespread deficits in critical thinking among Colombian university students, linking these gaps to poor self-regulation, weak metacognitive strategies, and difficulties with abstract thinking that often trace back to secondary schooling. Against this backdrop, the research team designed an ambitious experiment: could a didactic approach rooted in philosophy, applied to the notoriously abstract terrain of introductory quantum physics, move the needle on the cognitive skills most associated with academic survival and success?

The intervention centered on wave-particle duality, the foundational idea that entities such as electrons and photons behave as both waves and particles depending on how they are measured. This concept is an ideal testing ground for philosophical pedagogy because it directly violates classical intuitions about reality. The researchers structured their course around the Copenhagen interpretation, the most widely taught account of quantum mechanics, which treats the mathematical formalism as a tool for calculating probabilities of observations rather than a description of an observer-independent reality. Students were also exposed to alternative interpretations, including the hidden-variables approach of David Bohm, the objective-collapse model of Ghirardi, Rimini, and Weber, and the many-worlds interpretation of Hugh Everett—provocations designed to spark genuine debate rather than rote acceptance.

The didactic sequence unfolded over ten sessions of one hundred minutes each, spanning four weeks, and was implemented with two consecutive cohorts of engineering students in 2024: fourteen students in the first semester and eleven in the second, aged nineteen to twenty-two, all enrolled in a Waves and Modern Physics course for the first time. Each session combined scientific and philosophical readings drawn from accessible popular articles, videos, and simulators with the core activity of the approach: a Socratic dialogue in which the professor guided without imposing, using essential questions to expose contradictions and stimulate self-criticism. The design followed the philosophical didactics model of Sumachier, which organizes practice into three steps—asking questions to stimulate thinking, listening patiently, and problematizing situations that require reasoned decisions—echoing Matthew Lipman’s pioneering work on philosophy for children. Every session closed with a reflective summary and written responses to open-ended questions.

Assessment was anchored in Bloom’s taxonomy, which articulates critical thinking as six progressive cognitive levels: knowledge, comprehension, application, analysis, synthesis, and evaluation. Pilot testing confirmed what the literature predicted: lower-order skills could be fostered by conventional methods, but higher-order skills demanded more complex mediation and were conspicuously deficient. The team therefore narrowed its focus to the three highest levels—analysis, synthesis, and evaluation—designing a questionnaire of open-ended questions that was validated by seven expert judges in physics, mathematics, education, and engineering using Aiken’s V agreement statistics, ultimately yielding twenty-one validated questions, seven per skill. Students’ written answers were transcribed and coded in ATLAS.ti qualitative analysis software, with each response categorized as demonstrating either profound, significant, accurate reasoning or superficial, inaccurate reasoning. From these codings the researchers computed a polarisation ratio, an index ranging from minus one to plus one, where higher values indicate increasingly deep critical thinking, calculated session by session for each skill.

The results were remarkable for the analysis skill. In the first cohort, the polarisation ratio climbed from minus 0.55 to plus 0.41 points over the sequence; in the second, it rose from minus 0.75 to plus 0.45 points. Once improvement began in earnest from the fourth session onward, gains accrued at roughly 0.16 points per session in the first cohort and 0.20 points per session in the second. This meant that students who began the course producing inferences that contradicted the texts they read—confusing, for example, Einstein’s actual reaction to de Broglie’s thesis—ended it capable of nuanced abstraction. One student’s mid-course response captured the shift: the behavior of individual particles is probabilistic and indeterminate, while the propagation of the waves describing those probabilities is deterministic and law-governed, a duality across levels that the student correctly identified as a key idea of quantum mechanics. Co-occurrence analysis revealed that the analysis-of-principles code acted as the articulating hub connecting the other analytical sub-skills, underscoring that the ability to infer an author’s purpose and viewpoint drove much of the observed progress.

Synthesis skills also improved, though more modestly, rising by approximately 0.05 to 0.08 points per session. Early in the sequence, students struggled to combine ideas into coherent conclusions, but by the middle sessions some were proposing surprisingly sophisticated models of the double-slit experiment, imagining particles splitting into multiple trajectories that entangled to produce interference patterns. Graphical tasks—such as drawing how an electron might adapt its wave behavior to fit a stable orbit in a hydrogen atom—revealed partial but genuine conceptual construction, with classic misconceptions persisting alongside emerging quantum insights. The content analysis identified one especially instructive pattern: the ability to gather and communicate information functioned as a cross-cutting axis of synthetic thinking. Even students with limited expressive resources sometimes derived coherent predictions, suggesting that the reflective, reformulation-rich environment of Socratic dialogue allowed conceptually promising ideas to surface despite communicative limitations.

The evaluation skill, the most demanding tier of Bloom’s taxonomy, told a different story. Its polarisation ratio fluctuated erratically, ranging roughly between minus 0.33 and plus 0.17 points, with no clear trend in either cohort. Students frequently offered judgments based on superficial internal criteria or vaguely invoked theoretical concepts without genuine critique, and some fell into internal contradictions, endorsing an anti-realist view in one sentence while attributing measurement outcomes to the system itself in the next. The authors attribute this instability primarily to epistemological challenges: weaknesses in basic skills such as knowledge and application limited the foundation needed for rigorous evaluative reasoning, while students’ persistent classical intuitions and reliance on appeals to authority undermined independent judgment. The study’s conceptual-change framework, following Moreira and Greca, held that conceptions coexist rather than being replaced, and the data confirmed that early misconceptions continued to shape how students formulated critiques even as new quantum ideas were layered onto them.

The comparison between the two independent cohorts provided an important reliability check. Correlating the session-by-session polarisation ratios across cohorts revealed high agreement when the instrument was treated as a global construct, with analysis and synthesis exhibiting consistent patterns, while evaluation showed the expected instability. The researchers were candid about their limitations: the convenience samples were small, occasional student absences may have introduced fluctuations, and the professor’s initial inexperience in conducting Socratic debates sometimes allowed discussions to lose their philosophical depth, particularly affecting the practice of evaluative argumentation. Time pressure was also a real constraint, since the dialogues consumed hours that traditional courses devote to covering the syllabus, potentially squeezing the sessions in which synthesis is usually consolidated.

Nevertheless, the broader implications are difficult to ignore. At a moment when engineering programs across Latin America face dropout crises partly rooted in cognitive-skill deficits, this study demonstrates that philosophy need not be confined to humanities departments. The ten-session sequence, available online and deliberately free of heavy mathematical formalism, offers a flexible and replicable template that any instructor can adapt to university or upper-secondary contexts. The authors recommend that institutions build curricular flexibility into physics courses, incorporating the philosophy of science into educational objectives, and that teachers familiarize themselves with the logic of Socratic debate before deploying it. Future work will strengthen the assessment instrument through confirmatory factor analysis and add micro-lessons targeting the basic skills that philosophical dialogue alone does not cultivate. In an era defined by the second quantum revolution—quantum computing, quantum cryptography, and quantum sensing—the need for engineers who can think critically about foundational concepts has never been greater. This study suggests that a method of questioning older than Socrates himself may be one of the most modern tools available for building that capacity.

Subject of Research: The impact of a Socratic philosophical didactic approach on the development of critical thinking skills in engineering students learning quantum physics

Article Title: Impact of a philosophical didactic approach on the development of critical thinking: qualitative evidence from the teaching of quantum physics

Article References: Tuero-O´Donnell, J. D., Castrejón, G., & Olivo, P. G. (2026). Impact of a philosophical didactic approach on the development of critical thinking: qualitative evidence from the teaching of quantum physics. Journal of New Approaches in Educational Research, 15(1), Article 10. https://doi.org/10.1007/s44322-026-00059-y

Image Credits: AI Generated

DOI: 10.1007/s44322-026-00059-y

Keywords: critical thinking, quantum physics, Socratic dialogue, philosophical didactics, engineering education, Bloom's taxonomy, wave-particle duality, qualitative content analysis, higher education, conceptual change, Colombia, educational innovation

The legendary thought experiment proposed by John Archibald Wheeler in 1978, famously known as Wheeler’s delayed-choice experiment, delves deeply into these mysteries. Wheeler imagined an experimental setup where the decision to measure a photon’s particle or wave behavior occurs only after it has passed through the initial double-slit apparatus. This post-selection choice appears to retroactively determine the photon’s past, introducing a perplexing tension between quantum theory and classical notions of reality and causality.

Over the past decades, numerous experimental efforts have sought to realize Wheeler’s delayed-choice scheme, implementing variations such as beam splitters and quantum detectors to test whether the act of observation genuinely influences past events. These endeavors confirmed that the measurement choice, even when delayed, modulates the observed quantum behavior, reinforcing the non-classical correlation between observation and system state that transcends classical intuition.

Expanding upon Wheeler’s original scheme, a groundbreaking study from research teams at Ningbo University and the University of Science and Technology of China has introduced a novel conceptual and experimental innovation: the dual-selection delayed-choice experiment. Unlike prior implementations, which primarily control the presence or absence of the second beam splitter, this cutting-edge approach manipulates the insertion status of both the first and second beam splitters, effectively extending the delayed-choice paradigm to a richer and more nuanced quantum landscape.

This novel experimental architecture is realized by employing two ancilla qubits coupled through entanglement with the system qubit—representing the photon of interest in the interferometric setup. By harnessing entanglement and quantum control gates, such as controlled-NOT (CNOT) and controlled-Hadamard operations, researchers effectively simulate the presence or absence of beam splitters via quantum state manipulations rather than classical mechanical insertions. The first ancilla qubit encodes the insertion choice for the initial beam splitter through its measurement basis, intricately linked to the system qubit by a maximally entangled state. Simultaneously, the second ancilla qubit governs the state of a second controlled-Hadamard gate, quantifying the insertion of the second beam splitter.

An additional tunable phase shifter introduces a controlled relative phase, adding further versatility to the quantum interferometer. The carefully engineered combination of these elements simulates a dynamically adjustable interferometer, where both beam splitter choices are realized as quantum operations conditioned on entangled ancilla states. This represents a significant leap from traditional on/off mechanical configurations, offering unprecedented control over the measurement context within the quantum framework.

The experimental results derived from this dual-selection scheme compellingly demonstrate the wave-particle duality of photons with heightened diversity and complexity. Where conventional delayed-choice experiments elucidate a binary scenario—either a wave-like interference or particle-like path detection—this dual-selection approach reveals richer quantum behavior. The results echo the complementary principle postulated by Niels Bohr, reinforcing that the nature observed depends fundamentally on the experimental setup and measurement choices, even when these choices are embedded in entangled states and occur after the photon’s transit.

Fundamentally, this research underscores profound implications for the nature of quantum measurement, particularly around temporal ordering. Delayed-choice experiments challenge the assumption that cause precedes effect in straightforward ways. Here, the inseparability of measurement and system state suggests a more nuanced spatiotemporal correlation, intertwining measurement outcomes with choices that ostensibly take place "later" in time. The dual-selection approach provides a powerful platform to study these correlations in a controlled, tunable manner, allowing for further investigations into the fabric of quantum causality.

Beyond its foundational significance, this work also opens promising avenues for quantum technologies. The ability to control and manipulate measurement contexts via entangled ancilla qubits and quantum gates suggests new methods for designing quantum sensors, communication protocols, and computation schemes where the measurement basis and quantum control can be dynamically determined. This could enhance robustness, flexibility, and security in future quantum devices.

Moreover, the experimental infrastructure employed—a blend of quantum logic gates, entangled photons, and tunable phase control—demonstrates the sophistication achievable in state-of-the-art quantum optics laboratories. It embodies the convergence of abstract theoretical concepts with cutting-edge experimental physics, illustrating how quantum information science has become instrumental in probing the deepest questions of physics.

Looking forward, further research building on this dual-selection delayed-choice apparatus might venture into exploring multipartite entanglement scenarios, more complex interferometric networks, or even integration with matter-wave systems. Investigations into how such quantum delayed-choice configurations can be extended to massive particles or hybrid quantum systems can potentially unravel deeper layers of quantum measurement theory and possibly hint at new physics beyond conventional quantum mechanics.

This novel interferometer and dual-selection methodology thus signify not merely an experimental advancement but a conceptual advancement toward rethinking how measurement, observer choice, and quantum system histories interrelate. It compels the physics community to continue unraveling the rich tapestry connecting quantum phenomena, measurement back-action, and the fundamental nature of reality.

In essence, the experimental realization of Wheeler’s delayed-choice experiment with dual selections elevates the conversation on quantum measurement, complementarity, and causality to unprecedented heights. By encoding measurement choices into entangled ancilla states and leveraging quantum gates to simulate beam splitter insertions, researchers have crafted a versatile quantum playground. This playground invites deeper introspection into how we understand and interact with the quantum world—how observation shapes reality, even retroactively, and how the boundary between wave and particle remains elegantly, mysteriously fluid.

The implications ripple through quantum foundations and the practical realm alike, promising fertile ground for discoveries that might redefine the limits of quantum control and insight for years to come.

Subject of Research: Quantum wave-particle duality and delayed-choice experiments involving entangled photons and quantum control gates.

Article Title: Experimental realization of Wheeler’s delayed-choice experiment with dual selections

Web References: DOI: 10.1007/s11433-024-2587-y

Image Credits: ©Science China Press

Keywords

Quantum delayed-choice experiment, wave-particle duality, Wheeler’s experiment, dual-selection interferometer, quantum entanglement, controlled-Hadamard gate, controlled-NOT gate, quantum measurement, phase shift, quantum causality, quantum optics, quantum information