How to Pick a Science Research Topic & Idea: FULL GUIDE

Main Ideas

• A research idea is developed, not picked instantly. Finding a topic takes time, reading, note-taking, learning, and repeated refinement. Research is not a one-day project.
• Start with genuine curiosity, not buzzwords. Popular subjects such as AI, drones, or cryptocurrencies can be worthwhile, but they are not automatically better research topics. A less fashionable problem can still lead to significant work.
• Let interests emerge through exploration. Students who do not yet know what they want to study can begin broadly. Interests may change, so early exploration should not be restricted to one field.
• Research and engineering are related but distinct.
  • Engineering focuses on developing a solution to a problem.
  • Scientific research focuses on answering a question or testing what works.
  • A project may eventually involve both, but it helps to know which goal is driving it.

Suggested Process for Developing a Research Idea

1. Explore a wide range of material. Read books, blogs, articles, and introductory resources; watch videos and lectures. Look for clear, high-quality content, including free courses and educational materials. You do not need to choose a final topic right away.

2. Keep a regular record of what you learn. Use a notebook, journal, or digital document. Write down key points, unfamiliar terms, reactions, and questions prompted by each resource. Writing helps turn passive interest into active questions and ideas.

3. Follow promising topics further. If a video, article, or lecture catches your attention, investigate it more deeply. Wikipedia can be useful for orientation and for finding references, but it should not be treated as the final authority. Look up unfamiliar concepts using reference materials, search engines, or question-answering tools, then follow up with more substantive sources.

4. Move from general reading to scientific literature. Once a topic seems promising, read relevant scientific papers and review articles. Use them to understand what researchers already know, what methods they use, and what problems remain. Take notes and continue narrowing your focus rather than jumping straight to a project title.

5. Notice which topics keep drawing your attention. After exploring several areas, look for subjects you repeatedly return to or want to understand better. Narrow gradually—from broad subjects to subfields, then to specific issues or questions.

6. Turn an area of interest into a possible question or project. Ideas may start out rough; they do not need to be fully specified at first. Ask whether the goal is to answer a scientific question, develop an engineering solution, or combine the two. Identify the problem, possible approach, and evidence or evaluation that would show whether it works.

7. Evaluate and refine possible projects. Compare ideas by listing their advantages and challenges. For a student competition or research project, consider:

   • Is there enough time?
   • Is the project feasible?
   • Are the necessary materials, data, equipment, or expertise available?
   • What will it cost?
   • What skills will need to be learned?

Seek feedback when possible, especially before committing to a plan.

1. Learn the skills the project requires. Once the direction is clearer, identify gaps in your knowledge and learn what you need as you go. This may involve reading more papers, studying technical concepts, watching lectures, practicing programming, or learning new methods. Continue taking notes throughout the process.

2. Conduct the work, get feedback, and share the results. The presenter recommends doing the research first, then considering competitions for feedback and experience. Students can also share or publish results so others can learn from the work. He points viewers to separate resources on research opportunities, conducting research, competitions, science-fair projects, and publication.

Examples from the Presenter

• Pancreatic cancer project: The presenter’s interest grew from school biology, a visit to a lab focused on pancreatic cancer, and personal connections to the disease. He later combined that interest with an existing curiosity about AI. Reading about the disease and current methods led him toward a project using AI to improve radiation therapy for pancreatic cancer.
• Brain-tumor modeling project: In a research program and lab setting, he reviewed literature on brain tumors, including glioblastoma, and on how tumors are modeled and treated. He identified challenges and evaluation criteria, then developed a proposal involving patient MRI data and tumor-volume prediction. To pursue the idea, he learned additional methods, including neural ordinary differential equations. He says the resulting AI model aimed to predict future brain-tumor growth to help plan radiation and chemotherapy.
• Lesson from both examples: Neither project began as a fully formed idea. Each developed through prior interests, extensive reading, note-taking, identifying a problem, and learning the skills needed to investigate it.

Speakers and Sources Mentioned

• Speaker: Rishabh Jain, the presenter. The auto-generated captions render his name inconsistently.
• Educational and information sources: Kurzgesagt – In a Nutshell; Lex Fridman’s podcast; the Google blog; a Stanford natural-language-processing course on YouTube; Wikipedia; Google Search; ChatGPT; Sci-Hub; scientific articles and review papers; and NIPS conference videos.
• Research and competition organizations or publications: Nature Scientific Data, BMC Bioinformatics, Frontiers in Oncology, Regeneron ISEF, Broadcom MASTERS, the Junior Symposium in Science and Humanities, and an MIT/CEE summer research program.