Exercise: Think, don’t delegate

Author
Affiliation

Stefan Hackstein

Published

23 Jun, 2025

This activity investigates how we process information by probing various research modes and reflecting on how each affects comprehension and reasoning.

Learning Objectives

  • Strengthen knowledge of cognitive psychology principles and their application in learning and research.
  • Gain insight into how AI influences cognitive processes like critical thinking and comprehension.
  • Learn to evaluate when and how AI helps or hinders learning in educational and research contexts.

Task (15 minutes)

Exercise Flow

4 Research Questions → Choose 3 Questions → Apply 3 Different Research Modes
                                             ↓
                     Question 1 → Mode 1: Without Chatbot (5 min)
                     Question 2 → Mode 2: With Chatbot Only (5 min)  
                     Question 3 → Mode 3: With Chatbot Collaboration (5 min)
                                             ↓
                                    Record insights on Miro Board

In this exercise, we will explore three of the four questions below through different research modes, helping us experience the varying methods of inquiry while gaining valuable insights for the discussion that follows.

Important: You will use a different question for each research mode (3 questions total, one per mode). Choose any three questions from the list below, but ensure you follow the sequence of research modes in order.

  1. How does text-only learning shape comprehension, and what factors govern its effectiveness?
  2. How do brief summaries influence comprehension, and which elements drive their success?
  3. How does crafting a personal explanation boost critical thinking, and what are its limits?
  4. What makes partner feedback such an effective learning tool, and how can you use it best?
Task per Research Mode

For each research mode, you’ll work on a different question. Focus on the following:

  • How does the learner acquire knowledge using this approach?
  • When and how is that approach beneficial?
  • What are the pros and cons of this approach?

Add your answer as a sticky note in this Miro Board, within the frame for your question.

Research Modes

There are 3 sessions (5 minutes each) to approach the question differently:

  1. Without Chatbot: Research or discuss using classical sources or prior knowledge.
  2. With Chatbot Only: Use a single well-crafted prompt to get the best possible answer.
  3. With Chatbot Collaboration: Interact with the chatbot iteratively—ask follow-up questions, refine the answer collaboratively.

A short example will be shown for each mode before starting.

After engaging in fast-paced research and experiencing first hand how chatbots are usually applied in this context, we now want to reflect on the experience using the following steps:

  1. Think (1 minute): Take a moment to reflect individually on questions below.
  2. Pair (3 minutes): Discuss your reflections with a partner and compare your insights.
  3. Share (6 minutes): Share key takeaways with the group.

Questions to Consider

  1. How did the different sessions differ regarding how you learned and how you experienced that learning?
  2. What pros and cons did you experience using chatbots for learning?
  3. What’s your advice on when and how to use Chatbots for learning something new? Does it apply to novices and experts alike?

Key Takeaway

We explored how AI processes information and its impact on human learning. We generally find that using chatbots can reduce cognitive engagement. According to a recent study on the cognitive cost of using an LLM for essay writing (Kosmyna et al. 2025), this actually shows in lower brain activity, resulting in a lack of comprehension as well as in worse memorization of their own original text written on the topic.

The key take away is: When learning new concepts, use AI assistants strategically—to handle irrelevant administrative tasks and scaffold your thinking, not to bypass the productive struggle that builds genuine understanding. The cognitive effort you experience when grappling with new ideas isn’t a bug to be fixed; it’s the very mechanism through which expertise develops.

Back to top

References

Kosmyna, Nataliya, Eugene Hauptmann, Ye Tong Yuan, Jessica Situ, Xian-Hao Liao, Ashly Vivian Beresnitzky, Iris Braunstein, and Pattie Maes. 2025. “Your Brain on ChatGPT: Accumulation of Cognitive Debt When Using an AI Assistant for Essay Writing Task.” June 10, 2025. https://doi.org/10.48550/arXiv.2506.08872.

Reuse

Citation

BibTeX citation:
@online{hackstein2025,
  author = {Hackstein, Stefan},
  title = {Exercise: {Think,} Don’t Delegate},
  date = {2025-06-23},
  url = {https://virtuelleakademie.github.io/denken-statt-delegieren/exercises/},
  langid = {en}
}
For attribution, please cite this work as:
Hackstein, Stefan. 2025. “Exercise: Think, Don’t Delegate.” June 23, 2025. https://virtuelleakademie.github.io/denken-statt-delegieren/exercises/.