Eric Prenen pointed me to a recent Chinese study on adaptive teaching strategies in higher education. The authors, Panfeng Shi and Weijun Liu, developed a five-part approach for history classrooms, combining adaptive learning with positive psychology. The result? The experimental class improved from an average score of 68 to 86 in one semester. The control group even dropped slightly. The evidence seems clear.

And yet, something didn’t sit right. Everything fits — the theories (Bloom, Kolb, Dweck, Siemens), the design, the graphs, even a Hawthorne-effect control check. And still… something’s missing.

What stood out most was the lack of doubt. Dweck’s growth mindset is embraced as uncontested truth, without any mention of the ongoing debates about its long-term impact, context sensitivity, or limited effectiveness in real classrooms. Kolb’s learning cycle is treated as a blueprint for teaching design, yet its shaky track record when used to classify learners goes unmentioned.

Even learning styles creep back in, smuggled under the more palatable phrasing of “different learning styles, abilities and interests.” Sure, they probably just mean students are different and we should respond accordingly — fair enough. But why recycle terminology that has caused more confusion than clarity?

It’s a familiar pattern. Educational innovation often leans heavily on grand theories and the optimism of positive psychology, but forgets to ask whether the foundations still hold. The result is a stack of layered models and “integrated” strategies, all backed by a rising line on a chart.

And that’s where the real danger lies: the seduction of the graph. The comfort of instrumental thinking. If the numbers go up, it must work — even if we don’t know what’s improved or why. Even if we never ask students what moved them, confused them, or changed them.

I don’t blame the authors. Their intentions seem honest: more engaged students, better results. But education isn’t just about stacking interventions. Even the most well-meaning strategy risks becoming a neatly packaged illusion without critical reflection and room for uncertainty.

Or maybe, just maybe, I’m wrong — and this study challenges some of the myths we’ve discussed in our books, or at least adds nuance. But if that’s the case, we need more research to test it across different contexts.

Abstract of the study:

This work intends to meet China’s need for high-quality talents and optimize the current college classroom teaching modes. It first investigates the current situation of college history classroom teaching from the perspective of positive psychology and adaptive deep learning. Then, the work formulates a new teaching strategy. This teaching strategy is divided into five parts: verbal information, smart skills, cognitive strategies, action skills, and learning attitudes. Then, the proposed new teaching strategy is applied to practice. Sixty students from a university in Anyang are recruited and divided into Class A (experimental class) and Class B (control class). The average score of Class A using the proposed teaching strategy has increased by 18%, from 68 to 86 points. The average score of Class B without using the proposed teaching strategy has decreased by 1%, from 68 to 67 points. This indicates that the college history classroom-oriented teaching strategy based on adaptive deep learning is both scientifically sound and effective.