15.01Academic University ResearchUndergraduateAvailable

Academic Research Topic Selection

Turn an area of interest into one researchable question that fits your word count, your months, your access and your ethics timeline.

Free. Works on Claude, ChatGPT, Gemini or any assistant that accepts a skill file.

What this skill does

The method, encoded.

Most weak final-year projects are not badly executed. They are competently executed on a topic that could never have gone well, chosen at the moment of least knowledge and greatest anxiety.

This skill supplies the procedure that is usually missing. It separates an area from a topic, a topic from a problem and a problem from a question, then narrows on four axes (population, phenomenon, context, comparison) one move at a time, so the narrowing can be justified to a supervisor rather than merely performed. It tests answerability by asking you to write the finding you might get and then its opposite. It runs a targeted check on whether the question is already answered, without requiring a full literature review first.

It then screens feasibility on seven dimensions with fatal-flaw logic: data access, sample availability, ethics class and timeline, method skill, supervisor expertise, time costed by stage, and scope against word count. One red is disqualifying, and no amount of enthusiasm compensates.

It also settles the originality question. Undergraduate and honours originality is not doctoral originality, and the search for an idea nobody has ever had is the wrong search.

The skill interrogates, structures and screens. It does not write your project.

Best used for

  • Moving from an area of interest to a researchable question
  • Narrowing a topic that a supervisor has called too broad
  • Testing whether a topic can be done in the time and words available
  • Checking whether a question has already been answered before committing
  • Resolving originality paralysis at undergraduate and honours level
  • Choosing between several competing project ideas
  • Re-scoping a topic that has lost its data access or failed at proposal stage

Typical inputs

What you give it.

Area of interest in the student's own words, Degree level, discipline and artefact type, Word count and submission date, Whether primary data collection is permitted, Marking criteria and their weightings, Institutional ethics route and turnaround, Institutional AI use policy for the assessment, Optional supervisor research area, Optional pre-existing access to data, participants or a setting

Typical outputs

What you get back.

Constraint sheet with each item sourced and unknowns marked, Area to topic to problem to question ladder per candidate, Imagined finding and its opposite, per candidate, Targeted prior-answer check with verdict and items consulted, Contribution statement pitched at the degree level, Seven-dimension feasibility table with fatal-flaw scoring, Marking criterion fit note, Three-candidate comparison and recommendation, Dated commitment note with rejected candidates and reasons

Method coverage

What the skill works through.

  1. The difference between an area, a topic, a problem and a question
  2. Narrowing on four axes without over-narrowing
  3. Testing whether a question is answerable
  4. Checking whether your question has already been answered
  5. What originality means at undergraduate and honours level
  6. The seven-dimension feasibility screen
  7. Why access and ethics timelines decide more projects than interest does
  8. Matching a topic to the marking criteria
  9. Comparing three candidates instead of defending one
  10. Committing to a topic and stopping the search

Download

Free skill. One file.

Enter your email once. Every skill you download after that takes a single click.

How to install

Add the skill file and the five kernel protocols to a Claude Project, a ChatGPT Project, a Gemini Gem, or paste them at the top of any assistant conversation. Then give it your real research material, not a description of it.

Download skill

Questions

Common questions.

How do I know if my research topic is too broad?

Ask what an answer would look like. If any honest answer to your question would be a summary of what is already known rather than a finding, the topic is too broad. The fix is to narrow on one axis at a time: who exactly, what precisely, in what setting, compared to what or when. Narrowing all four at once usually leaves a question so specific that no literature speaks to it.

What is the difference between a research topic and a research question?

A topic is a bounded subject. A problem is a tension inside it, stated as a sentence with a "but" or a "yet". A question is an interrogative sentence naming a population, a phenomenon and a context, whose answer would resolve part of that problem. Most students have a topic and believe they have a question, and the missing level is almost always the problem.

Does an undergraduate dissertation have to be original?

Not in the doctoral sense of contributing new knowledge to the field. At undergraduate and honours level the requirements are that the work is your own, that the question is not trivially answered, and that you demonstrate competent research practice. A known question examined in a new population or setting, a replication where the original may not hold, or an established framework applied to an unstudied case all qualify.

How do I check whether my topic has already been done?

Three targeted probes, before any full review: search your question in the vocabulary of the field and read abstracts only; find one or two recent review articles and read their future-research sections, which are the most efficient statement of what remains open; and check whether the question has been answered in a different population, which often converts an unoriginal question into a defensible one. A contested question, where published work disagrees, is usually the strongest position for an undergraduate project.

How long does a research project actually take?

Cost it by stage with dates rather than as a total: ethics approval, recruitment, data collection, transcription or cleaning, analysis, writing, revision. Ethics approval is the stage most often underestimated and the one most likely to make a topic infeasible, because it happens before anything else can start.

My supervisor says my topic is too big but not how to fix it. What do I do?

Write the question out with its four axes visible, then narrow the one that is doing the least work for your argument. Usually that is the population or the setting. Bring the supervisor two narrowed versions rather than asking again, because the conversation is much faster when there is something specific to react to.

Should I choose the topic I find most interesting?

Among topics that pass a feasibility screen, yes, because interest is what gets a project finished. Among topics that do not, no. Projects fail on data access far more often than on boredom, and a confirmed route to data is not a compromise, it is the difference between a project that exists and one that does not.

Can AI choose my dissertation topic for me?

It should not, and in many institutions it may not. AI assistance can generate candidates to react to, apply a feasibility screen consistently, and ask the questions a supervisor would ask. The choice, the justification and the words must be yours, you must be able to defend the topic unaided, and you must follow your institution's AI use and declaration policy for the assessment.

Research where people already are.
Analyse it where you already work.

Yazi helps researchers conduct surveys, AI interviews and longitudinal research directly through WhatsApp.

New Report on SA Gambling Impact
Check It Out