█
LastWrite
  • > Curriculum
  • > Pricing
  • > For Educators
  • > About
  • > Contact
Log InGet Started

Questions, concerns, bug reports, or suggestions? We read every message, write to us at [email protected].

More ways to reach us →
LastWrite

Structured computer science lessons for aspiring developers and security professionals.

[email protected]

(201) 785-7951

Mon–Fri, 9 AM–5 PM EST

Learn

  • Curriculum
  • Pricing

Company

  • About
  • For Educators & Schools
  • Contact Us

Legal

  • Terms of Service
  • Privacy Policy
© 2026 LastWrite. All rights reserved.
Curriculum/LLM Research and NLP/Research Practice and Publishing at ACL SRW/Scoping a Research Question
40 minIntermediate

Scoping a Research Question

After this lesson, you will be able to: Scope a research question that is novel enough, feasible on your budget, and the right size for an SRW paper, and find a mentor.

The hardest research skill is picking the right question. Too big and you cannot finish; too small and it is not a contribution. This lesson teaches scoping for an SRW-sized result and how to get mentorship.

Prerequisites:Reading Papers and Literature Review

What makes a good SRW question

Three tests. Novelty: it is not already answered (your literature review tells you). Feasibility: you can actually run it on free/cheap GPUs in a few months. Interest: someone in the field would want to know the answer. SRW papers are short, so the right size is one sharp question, not a system. Good shapes: 'Does method X, claimed for task A, also help task B?' 'Is the reported gain of X robust to Y?' 'A simple baseline matches X at a fraction of the cost.' 'Here is a failure mode of X and why.' Small and sharp beats broad and shallow.

💡 Reproduction is the safest first question

For a first paper, reproducing and extending an existing result is the highest-success path. You inherit a well-defined problem, baselines, and metrics, and your contribution is the careful reproduction plus one twist (a new dataset, an ablation the original lacked, a robustness check, a cheaper method that matches it). It sidesteps the riskiest part (inventing a problem) while still being a real contribution. The next lesson is dedicated to it.

Finding a mentor

You do not have to do this alone, and at SRW you are not expected to. Paths: a professor or grad student at your institution (even cold-emailing one whose paper you reproduced, with a specific, prepared question, works more often than students expect), the SRW mentorship program itself (accepted authors get a senior mentor; some years offer pre-submission mentorship), online communities, and your reproduction's original authors (researchers often respond to a thoughtful, specific question about their work). A mentor mostly helps you scope and avoid known pitfalls, which is exactly where first-time researchers struggle.

Common mistakes only experienced researchers catch

Picking a question so broad it cannot finish (the classic first-paper failure). Picking one so narrow it is not a contribution. Choosing a topic with no feasible evaluation on your hardware. Not checking novelty thoroughly, then discovering the question was answered last year. Trying to invent a brand-new method instead of doing a clean reproduction-plus-extension. Working in total isolation when a 20-minute mentor conversation would have caught a fatal flaw early.

Quick Check

Why is reproducing-and-extending an existing result the recommended first SRW project?

Pick the best reason.

Sign in and purchase access to unlock this lesson.

Sign in to purchase
←Reading Papers and Literature Review
Back to Research Practice and Publishing at ACL SRW
Reproducing a Paper→