How to Use AI to Help You Dig Deep into Any Research Topic

Back to AI Tips, Tutorials, & Recordings

Introduction

I put this guide together because I’ve often found myself stuck at the beginning of a research project—not knowing where to start, how to organize my thoughts, or which sources to trust. It’s overwhelming to sift through endless papers trying to gather relevant sources and understanding methodologies. This guide aims to give structured prompts that urge the AI tool to evaluate and craft source lists, explain methods and summarize intelligently – making the research process efficient and manageable.

Specific Examples

Note: Follow guidelines set by textbook and journal publishers for the allowable and prohibited uses when uploading copyrighted materials to AI tools per Standards and Recommendations for the Use of Generative AI in Teaching and Learning at Northeastern.

General Research Question
- Quick, foundational prompts to understand a complex topic or issue by breaking it down into clear parts.
Compare Perspectives on an Issue
- Prompts designed to explore opposing or varied viewpoints on a debate, theory, or controversial claim.
Summarize a Research Paper or Article
- Focused prompts to extract the purpose, methods, findings, and implications of a single academic or journalistic source.
Deep Research Investigation
- In-depth prompts that unpack a multi-dimensional issue across historical, technical, ethical, and future-facing angles.

Tools and Resources

Claude (with Deep Research Mode) – Excellent for in-depth academic exploration. Use Deep Research Mode to synthesize complex sources, extract key insights from long documents, and generate thoughtful, structured summaries or follow-up questions. Great for tackling dense readings or prepping study notes.
ChatGPT – Helpful for summarizing material, generating ideas, and turning research into outlines, guides, or practice questions.
Perplexity.ai – Best for quick answers linked to current sources, with citations and follow-up suggestions.
Semantic Scholar – An AI-enhanced academic search engine with tools for deep research, including citation graphs and topic mapping.

Prompts

Each template walks the AI through logical steps for better quality and source-backed answers. Cut and paste all of the text in each section into the chat field and customize by replacing the content in brackets.

General Research Question

“You are a research assistant helping a student understand a topic. The student asks about: [Insert main topic or research question]”

Steps:

Break Down the Main Question into Sub-Questions if Needed
- Clarify what the student is really asking.
- Identify any sub-questions or related themes that help unpack the main topic.
- Consider aspects such as causes, effects, stakeholders, trends, historical context, and current relevance.
Search and Summarize Credible, Recent Sources on the Topic
- Identify scholarly articles, government reports, reputable news sources, or authoritative databases.
- Summarize the most relevant and recent findings and perspectives.
- Aim for balance—include differing viewpoints or schools of thought if applicable.
Provide a Clear Summary in 2–3 Paragraphs
- Write a concise, informative explanation suitable for a student audience.
- Integrate key ideas from the sources without copying or overly quoting.
- Emphasize clarity, neutrality, and relevance.
Cite 3–5 Reputable Sources (with Links if Possible)
- List each source used, including author, title, and publication date.
- Provide links to the original articles or documents when available.
- Ensure sources are academically sound or widely recognized as credible.
Point Out Any Areas of Debate, Uncertainty, or Gaps
- Highlight unresolved questions or ongoing controversies.
- Note if data is limited, conflicting, or outdated in certain areas.
- Mention any emerging trends or areas that need further research.

Follow-Up Questions:

“What are the current trends or changes in this field?”
“What do experts disagree on, and why?”
“Are there any real-world applications or consequences?”
“What’s missing from the research?”

Compare Perspectives on an Issue

“Topic to compare: [Insert issue, debate, or controversial claim]”

Steps:

Identify the Central Issue and the Main Perspectives
- Clearly define the core issue or controversy at stake.
- Identify the main perspectives or positions that exist.
- Note if these perspectives come from specific disciplines, political ideologies, communities, or schools of thought.
Summarize Each Side’s Key Argument and Who Supports It
- For each perspective, outline its central claims or rationale.
- Identify the individuals, organizations, institutions, or thought leaders who advocate for each side.
- Indicate whether each argument is based on ethics, empirical evidence, historical precedent, economic logic, etc.
Provide Examples or Studies Backing Each Position
- Cite specific studies, data, or real-world examples that each side uses to support its stance.
- Highlight methodological strengths or limitations of these examples, if relevant.
- Ensure each side is represented with balanced evidence, not just opinions.
Include Citations or Links for Verification
- Provide full citations, hyperlinks, or references to source material mentioned.
- Use reliable academic, journalistic, or institutional sources.
- When quoting, indicate authors and publication dates where possible.
Conclude with a Neutral Summary of How the Views Align or Diverge
- Compare where the perspectives overlap, and where they fundamentally differ.
- Avoid taking a stance—this is a synthesis, not an argument.
- End with an open-ended question or note about unresolved areas, emerging perspectives, or future implications.

Follow-Up Questions:

“Which side has stronger empirical support?”
“Are there any middle-ground or hybrid perspectives?”
“How has this debate evolved over time?”
“What might shift public or expert opinion in the future?”

Summarize a Research Paper or Article

“Article/paper to summarize: [Paste link or full text]”

Steps:

Identify the Title, Author(s), and Purpose
- Explain the purpose of the article: What is it aiming to contribute? Is it theoretical, empirical, applied, or a review?
Explain the Research Question or Hypothesis
- What is the central question or problem the article addresses?
- If applicable, state the hypothesis being tested.
- Explain the relevance of the research question to the broader field.
Summarize the Methodology
- Describe the research design (qualitative, quantitative, mixed methods, etc.).
- Identify the subjects, dataset, or materials used.
- Outline how data was collected (e.g., surveys, experiments, interviews, archival analysis).
- Describe the methods of analysis or tools employed (e.g., statistical tests, coding schemes, software).
State the Main Findings
- Summarize the key results and whether they support the hypothesis or answer the research question.
- Mention any significant patterns, trends, or relationships uncovered.
Share the Conclusions and Implications
- What conclusions do the authors draw from their findings?
- How do the results impact the field, policy, theory, or practice?
Mention Any Limitations or Disclosures
- Identify any weaknesses in the study’s design, sample, or generalizability.
- List any author disclosures, such as funding sources or potential conflicts of interest.
- Note whether the authors acknowledge these limitations or suggest improvements.

Follow-Up Questions:

“How do these findings compare with similar studies?”
“What are the limitations and how could future research address them?”
“What assumptions does the study make?”
“Who might benefit from or be affected by this research?”

Deep research Investigation Prompt

“You are a research assistant conducting a deep investigation into the following topic:[Insert complex topic, issue, or emerging field]”

Steps:

Clarify the scope of the topic: Define key terms, outline sub-domains, and identify why this topic matters (social, technological, ethical, economic relevance).
Divide the topic into 3–5 dimensions or sub-questions (e.g., historical background, current state, competing approaches, future directions, policy implications).
For each dimension, collect insights from at least 2–3 credible sources (e.g., peer-reviewed papers, think tanks, whitepapers, major media). Summarize each dimension in 1–2 detailed paragraphs.
Identify points of consensus and disagreement among experts, and explain the reasons for any controversy or uncertainty.
Provide a high-level synthesis tying the dimensions together into a full picture. Explain how this understanding might inform decisions, innovation, or research.
Include a reference list of at least 5–7 reputable sources with links, publication names, and dates.
End with 2–3 open research questions or areas for exploration that remain unresolved.

Follow-Up Questions:

“What are the most credible institutions or experts leading this research?”

“How is the conversation around this topic shifting over time?”

“What real-world impact does this have now — or could it have in 5–10 years?”

“How do different cultures, disciplines, or industries approach this issue?”

Thoughts and Concerns to Keep in Mind

Transparency of Prompts and Process: Document how AI was used—include your prompts, what versions/models were involved, and at what stages in the research process. This helps with reproducibility and ethical review.
Model Limitations: Be aware of the AI’s training cutoff (e.g., knowledge up to 2024/2023) and lack of access to real-time data unless explicitly updated. It doesn’t “know”—it predicts based on patterns.
Ethical Use & Plagiarism: Don’t copy AI-generated content into academic work without rewriting and checking originality. Treat it like a brainstorming assistant, not a ghostwriter.
Underrepresented Voices: Proactively ask AI to surface less dominant perspectives, especially from marginalized communities or global viewpoints often absent in mainstream training data.
Hallucinations & Confabulations: AI can confidently “invent” facts, quotes, or sources. Always fact-check, especially with niche or historical data.
AI’s Role in the Workflow: Clarify whether AI is being used for ideation, summarization, outlining, hypothesis generation, etc.—and be transparent about that in methodology sections.

Back to AI Tips, Tutorials, & Recordings