This file compiles research that supports the main claims and practices in Chapter 6. It’s organized so you can quickly pull a justification, a citation, or a manuscript-ready line.
Research map (quick scan)
Instant answers and cognitive offloading (the “Google effect”)
AI assistance and engagement/ownership during writing tasks
Curiosity as a learning-and-memory booster (dopamine/hippocampus)
Curiosity and patience for discovery (process over closure)
Question-asking declines in school - but question-asking practice helps
Boredom supports creativity (when it’s not interrupted)
Observation-first instruction (explore-before-explain)
Verification and hallucinations (why “How would we check that?” matters)
Practical inquiry protocols (QFT; See-Think-Wonder)
- Instant answers, cognitive offloading, and memory
Core idea
When people expect that information will be available later, they’re more likely to offload memory - remembering where information is rather than what it is. This is often discussed as transactive memory or “cognitive offloading,” and the popular label is the “Google effect.”
What it supports in Chapter 6
Your claim that instant answers can train “relief” more than “understanding,” and why kids need practice staying with a question before closing it.
Manuscript-ready phrasing (feel free to paste)
When information is always one click away, the brain starts saving energy. It remembers where to find answers more than it remembers the answers themselves - which means kids get fewer reps at understanding.
Sources
[1], plus synthesis context in [11] and [12].
- AI assistance, engagement, and “ownership” in writing
Core idea
A controlled essay-writing study compared brain-only, search-engine, and LLM-assisted writing. The LLM-assisted group showed the lowest engagement on EEG measures and reported the lowest ownership of what they produced.
What it supports in Chapter 6
Your “tool ends the work vs creates the next step” rule, and the signs that the tool is doing too much (polished output with low understanding).
Manuscript-ready phrasing (feel free to paste)
If the tool does the first move every time, kids can look finished without becoming capable. The output improves. The ownership shrinks.
Sources
[2].
- Curiosity boosts learning and memory (it’s biology)
Core idea
Curiosity is associated with increased learning and memory for the answer - and often for incidental information encountered while curious. Research links curiosity to interactions between reward circuitry and hippocampus-dependent learning.
What it supports in Chapter 6
Your argument that the “wonder gap” before an answer arrives matters, and that curiosity is a practice worth protecting.
Manuscript-ready phrasing (feel free to paste)
Curiosity isn’t fluff. It’s a brain state that makes learning stickier - which is why instant closure can be a surprisingly expensive habit.
Sources
[5] (primary), with an accessible summary in [13].
- Curiosity can increase patience for discovery
Core idea
Research on dynamic information gathering suggests curiosity can promote willingness to let information unfold rather than shortcutting to immediate resolution. The process of discovery has value, not just the answer.
What it supports in Chapter 6
Your push against premature closure: “teach children the skill of not being finished too early.”
Manuscript-ready phrasing (feel free to paste)
A good question can make a kid slower, not faster - because the process becomes part of the reward.
Sources
[14], plus a plain-language summary in [15].
- Question-asking declines in school - but it’s trainable
Core idea
Susan Engel’s work argues that curiosity is intrinsic but often not cultivated in typical classrooms; question-asking can drop sharply in school contexts. More recent experimental work shows that practicing question-asking can increase curiosity and improve science learning in young children (especially for kids with less background knowledge).
What it supports in Chapter 6
Your Wonder Wall, the Question Upgrade ladder, and the idea that “better questions” can be taught like any other skill.
Manuscript-ready phrasing (feel free to paste)
Kids don’t stop asking questions because they’re lazy. They stop because the room rewards being done. Change the rewards, and you can get the questions back.
Sources
[3], [4], plus the shorter overview in [16].
- Boredom supports creativity (when we don’t interrupt it)
Core idea
Studies suggest that boredom can increase creative idea generation afterward. A common explanation is that boredom nudges the mind into daydreaming/mental wandering, which supports novel connections.
What it supports in Chapter 6
Your claim that boredom is “the doorway,” and your three-minute pause rule.
Manuscript-ready phrasing (feel free to paste)
Boredom is the preheating stage for imagination. If you interrupt it instantly, the oven never gets warm.
Sources
[6], plus additional popular summaries in [17] and [18].
- Observation-first learning: explore before explain
Core idea
Explore-before-explain approaches prioritize firsthand experiences and meaning-making before formal explanation, especially in science learning. This tends to increase engagement and understanding because students build questions from real contact with reality.
What it supports in Chapter 6
Your sequence: observation first, questions second, tools third.
Manuscript-ready phrasing (feel free to paste)
Let kids touch reality before they download a summary of it.
Sources
[19], [20].
- Verification and hallucinations: why “How would we check that?” matters
Core idea
LLMs can produce confident but incorrect information (often called hallucinations). Verification is a key literacy practice - not cynicism - because fluency is not the same as truth.
What it supports in Chapter 6
Your verification habit and the road-crossing metaphor.
Manuscript-ready phrasing (feel free to paste)
Treat AI like an eager intern: helpful, fast, and sometimes confidently wrong. Teach kids to verify without drama.
Sources
[9], [10], plus OpenAI’s deeper explanation in [21].
- Practical inquiry protocols (classroom and home)
Core idea
QFT (Question Formulation Technique) is a structured protocol for generating, improving, and prioritizing questions. Project Zero’s “See, Think, Wonder” is a short routine that scaffolds observation, interpretation, and wonder. Both map neatly onto the chapter’s “observation-first” theme.
What it supports in Chapter 6
Your prompt pack and Question Upgrade Ritual; using AI as a tutor or coach rather than an answer vending machine.
Manuscript-ready phrasing (feel free to paste)
Question-making is a skill set. When you teach it explicitly, kids stop asking for answers and start designing investigations.
Sources
[7], [8].
Full reference list (numbered)
[1] Sparrow, B., Liu, J., & Wegner, D. M. (2011). Google Effects on Memory: Cognitive Consequences of Having Information at Our Fingertips. Science, 333(6043), 776-778. https://www.science.org/doi/10.1126/science.1207745
[2] Kosmyna, N., Hauptmann, E., Yuan, Y. T., Situ, J., Liao, X.-H., Beresnitzky, A. V., Braunstein, I., & Maes, P. (2025). Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task. arXiv:2506.08872. https://arxiv.org/abs/2506.08872
[3] Engel, S. (2011). Children’s Need to Know: Curiosity in Schools. Harvard Educational Review, 81(4), 625-645. https://elimindset.com/wp-content/uploads/2022/08/engel2011.pdf
[4] Park, A. T., Colantonio, J., et al. (2025). Question asking practice fosters aspects of curiosity in science content in young children. npj Science of Learning. https://www.nature.com/articles/s41539-025-00384-5
[5] Gruber, M. J., Gelman, B. D., & Ranganath, C. (2014). States of Curiosity Modulate Hippocampus-Dependent Learning via the Dopaminergic Circuit. Neuron, 84(2), 486-496. https://www.sciencedirect.com/science/article/pii/S0896627314008046
[6] Mann, S., & Cadman, R. (2014). Does Being Bored Make Us More Creative? Creativity Research Journal, 26(2), 165-173. https://www.tandfonline.com/doi/abs/10.1080/10400419.2014.901073
[7] Right Question Institute. (n.d.). What is the Question Formulation Technique (QFT)? https://rightquestion.org/what-is-the-qft/
[8] Project Zero, Harvard Graduate School of Education. (n.d.). See, Think, Wonder. https://pz.harvard.edu/resources/see-think-wonder
[9] OpenAI Help Center. (n.d.). Does ChatGPT tell the truth? https://help.openai.com/en/articles/8313428-does-chatgpt-tell-the-truth
[10] Huang, L., et al. (2023). A Survey on Hallucination in Large Language Models. arXiv:2311.05232. https://arxiv.org/abs/2311.05232
[11] Gong, C., et al. (2024). Google effects on memory: a meta-analytical review. https://pmc.ncbi.nlm.nih.gov/articles/PMC10830778/
[12] Firth, J., et al. (2019). The “online brain”: how the Internet may be changing our cognition. World Psychiatry. https://pmc.ncbi.nlm.nih.gov/articles/PMC6502424/
[13] University of California (UC Davis / UC). (2014). Curiosity helps learning and memory (news summary). https://www.universityofcalifornia.edu/news/curiosity-helps-learning-and-memory
[14] Hsiung, A., et al. (2023). Curiosity evolves as information unfolds. PNAS, 120(42), e2301974120. https://www.pnas.org/doi/10.1073/pnas.2301974120
[15] Scientific American (2024). No Spoilers, Please! Why Curiosity Makes Us Patient. https://www.scientificamerican.com/article/no-spoilers-please-why-curiosity-makes-us-patient/
[16] Engel, S. (2021). Why Do Kids Ask Fewer Questions When They Start School? TIME. https://time.com/5941608/schools-questions-fostering-curiousity/
[17] Burkus, D. (2014). The Creative Benefits of Boredom. Harvard Business Review. https://hbr.org/2014/09/the-creative-benefits-of-boredom
[18] TIME (2019). Being Bored Can Be Good for You - If You Do It Right. https://time.com/5480002/benefits-of-boredom/
[19] Brown, P. (2023). Exploring Before Explaining Sparks Learning. Edutopia. https://www.edutopia.org/article/explore-before-explain-elementary-science/
[20] NSTA (2019). Explore Before Explain (resource PDF). https://static.nsta.org/pdfs/201910BookBeatLearningAboutEcosystems.pdf
[21] OpenAI (2025). Why language models hallucinate. https://openai.com/index/why-language-models-hallucinate/