Critical Thinking on Autopilot? Prompts to Deepen Your Discussion Section

If Chapter 4 (Results) is where you show the examiner what you found, Chapter 5 (Discussion) is where you must explain why it matters.

This is the chapter where most postgraduate students at universities like UM, UKM, and UTM stumble. The most common feedback from examiners during a Viva Voce is: “Your discussion is too descriptive. You are simply repeating your results.”

Writing a high-level Discussion chapter requires synthesizing your raw data with the theoretical framework you established in your Chapter 2. When panicked students feed their results into an LLM and type, “Write my discussion chapter,” the AI responds with generic and hallucinated fluff filled with words like “pivotal,” “underscores,” and “delves.” Not only does this fail to demonstrate critical thinking, but it also guarantees a massive red flag on the Turnitin AI Checker.

You cannot put critical thinking on autopilot. However, you can use Academic Prompt Engineering to turn an LLM into a world-class sounding board. By forcing the AI to generate hypotheses, compare literature, and frame limitations, you can avoid writer’s block and enhance all your academic arguments.

Here are the exact prompts to improve your Chapter 5 from a simple summary to a strong academic critique.

1. The Literature Comparator (Connecting the Dots)

Your findings do not exist in a vacuum. You must explicitly connect your data to the existing literature. Do your results align with previous studies, or do they challenge the current consensus?

The Prompt:

“Act as an expert peer reviewer in the field of [Insert Field]. My recent study found that [Insert Key Finding]. In Chapter 2, I reviewed Smith (2024), who found [Finding A], and Lee (2023), who found [Finding B]. Write a 250-word synthesis paragraph that compares my finding with Smith and Lee. Do not just summarize; explicitly analyze why my findings might align with one and diverge from the other based on our differing contexts [e.g., Malaysia vs. Europe]. Maintain a highly critical, objective academic tone.”

Why this works: It prevents the AI from hallucinating fake citations by forcing it to use the exact authors you provide. It also shifts the focus from “summarizing” to “comparing,” which is the core of high-level academic writing.

2. The Contradiction Explainer (Handling Unexpected Results)

Unexpected results are not a failure, but rather are an opportunity for a good discussion. If your data contradicts established theories, then examiners will wait to see how you defend it. The AI can help you brainstorm academic justifications for these anomalies.

The Prompt:

“My empirical data revealed a surprising result: [Insert Unexpected Finding]. This directly contradicts the established [Insert Theory/Author]. Generate a list of 4 highly academic, plausible hypotheses that explain this discrepancy. Consider variables such as methodological differences, sample size limitations, cultural/geographic context, or changes in the industry landscape since the original theory was published. Give me the hypotheses only; do not write the final paragraph.”

Why this works: By asking for hypotheses rather than a finished paragraph, you use the AI as a brainstorming partner. You can then select the most logical explanation, investigate it, and write the final argument in your own voice.

3. The Theoretical Framework Anchor

A common structural failure in thesis writing is the “abandoned theory.” You spent 10,000 words in Chapter 2 explaining a theoretical framework (e.g., Transformational Leadership Theory or Technology Acceptance Model), but you never mention it again in Chapter 5. You must anchor your new data back to that theory.

The Prompt:

“I am using the [Insert Name of Theoretical Framework] to ground my research. My primary empirical finding is [Insert Finding]. Write a rigorous academic paragraph explaining how my specific finding supports, extends, or challenges the core tenets of this theory. Use the specific terminology associated with the theory (e.g., [Insert 2-3 key terms from the theory]). Do not use generic transition words.”

Why this works: Supplying the AI with the specific terminology forces it to write with disciplinary precision, avoiding the “average internet voice” that Turnitin easily detects as AI-generated.

4. The Limitation Framer (Protecting Your Viva)

Every study has flaws. If you do not point them out, your examiner will. Writing the “Limitations of the Study” section is about acknowledging boundaries without destroying the credibility of your entire thesis.

The Prompt:

“I need to write the ‘Limitations and Future Research’ section for my Chapter 5. My study had the following limitations: [List 2-3 limits, e.g., a cross-sectional design that prevents causal claims, and a small sample size restricted to Kuala Lumpur]. Write a paragraph framing these limitations constructively. Explain why these constraints were necessary or unavoidable. Then, propose two highly specific, actionable directions for future researchers to address these exact gaps.”

Why this works: It teaches the AI the art of academic diplomacy. It frames your limitations as accepted boundaries of the research scope rather than fatal methodological errors.

Humanizing Your Discussion for Turnitin

Because Chapter 5 requires heavy interpretation and transition-heavy phrasing, it is the chapter most vulnerable to high AI similarity scores. Turnitin’s 2026 algorithm specifically hunts for the exact type of “analytical” language default LLMs produce (e.g., “This study illuminates the multifaceted nature of…”).

To ensure your discussion chapter passes the IPS formatting and Turnitin checks:

1. Inject Your Voice: Use the AI’s structural suggestions, but rewrite the core arguments using your own vocabulary and writing style.
2. Delete the Drama: AI loves to exaggerate findings. Remove words like groundbreaking, massive, pivotal, or shocking. Replace them with significant, substantial, or notable.
3. Use a Human Editor: A critical eye is required to ensure your arguments actually make logical sense.

If you have drafted your Chapter 5 using these prompts but it still feels disjointed or robotic, our Academic Editing and AI Humanizing Services can help. We tighten your arguments, strip out the AI jargon, and ensure your manuscript reads like the work of a seasoned academic.

Further Reading: The Researcher’s Prompt Library: Tested LLM Prompts for Thesis Writing & Journal Publishing