Prompt Design: Simple vs. Structured (AI “Brain Rot” Research Example) Simple Prompt find a topic for research in brain rot related to ai Structured Prompt Role: You are a senior research advisor in cognitive science + media studies. Task: Generate 8-12 specific, novel, and feasible research topics on “brain rot” related to AI. By “brain rot,” treat it as a popular term capturing potential cognitive/affective changes (e.g., attentional fragmentation, reduced deep processing, reward sensitivity) caused by AI-curated or AI-generated content. For each topic, deliver: • Title ( ≤ 12 words) • One-sentence gist • Core research question (RQ) • Why it matters (identify the gap) • Operational definition of “brain rot” for this topic (what constructs, how measured) • Methods & design (quant/qual/mixed; cross-sectional/longitudinal/experiment) • Sample & setting (target N, demographics, inclusion) • Data sources (public datasets OR collection plan) • Analysis plan (stats or qualitative coding approach) • Risks/ethics (privacy, algorithmic exposure, minors) • Feasibility rating (time/cost/IRB complexity: Low/Med/High) with 1-2 lines justification • Novelty rating (1-5) with a one-line rationale Constraints to respect: • Audience level: [undergrad honors / master’s thesis / PhD qualifying paper] = [choose one] • Time budget: [4-6 months] • Max N: [ ≤ 300] if collecting new data • Include at least 1 neuroscience angle AND 1 education-focused study • Platforms to prioritize: TikTok/YouTube/Instagram + generative AI (e.g., ChatGPT) • Prefer outcomes: attention, working memory, deep reading, metacognition, mood/irritability Output format: 1) A ranked table (Rank, Title, RQ, Methods, Feasibility, Novelty) 2) 150-200 n word mini n abstracts for the top 3 topics 3) A 6-8 item reading starter pack (2018-present, ≥ 2 from 2023+), include DOI/URL and 1-2 sentence takeaway for each citation. Why the Structured Prompt Wins Dimension Simple Prompt Structured Prompt Goal clarity Vague: find a topic Explicit: 8–12 topics with scope and constraints Role & context None Sets expert role + domain framing Definitions Assumes meaning of “brain rot” Operationalizes constructs & measures Deliverables Unspecified Ranked table + abstracts + reading pack Methods guidance None Specifies methods, samples, analysis plans Feasibility guardrails None Time budget, max N, ethics/risks Coverage constraints None Neuro + education angles; platforms; outcomes Before/After Example Outputs Simple Prompt Output: • "Impact of AI on attention" • "Does TikTok cause brain rot?" Structured Prompt Output: • "Short n form AI n curated feeds and sustained attention: 4 n week field experiment in Indian undergrads (N ≤ 300)" • "ChatGPT n assisted study plans and deep reading in first n year engineering courses: randomized classroom trial" • "Dopaminergic reward sensitivity to AI n generated meme streams: EEG oddball task with pupilometry" Pro Tips for Writing Structured Prompts • State a role that implies standards or domain expertise. • Define key terms operationally (constructs, measures, time horizon). • Constrain deliverables (count, format, fields, length limits). • Bound feasibility (time, sample size, data access, ethics). • Name required angles, populations, platforms, and outcomes. • Ask for ranking or prioritization to force trade n offs. Quick Links to Major LLMs Perplexity — https://www.perplexity.ai ChatGPT — https://chat.openai.com Claude — https://claude.ai Grok — https://x.ai Summary Insight A structured prompt adds context, clarity, and constraints—guiding AI to deliver high-quality, actionable, and domain-specific responses.