A journalism student sits down to write a proposal and stares at a blank page for longer than they'd like to admit. They know they care about misinformation, or about how audiences behave online, or about the pressures newsrooms face under shrinking budgets. What they don't have is a question. And that gap between caring about a subject and being able to interrogate it is where most research proposals quietly fail before they've even begun.

This is not a problem of imagination. Most people drawn to journalism and media studies already notice things worth investigating a strange pattern in how a story spread, a contradiction between what a newsroom claims and what it publishes, a shift in how young audiences consume news compared to their parents. The difficulty is converting that noticing into something that can actually be studied: something with edges, a method, and an answer that isn't already obvious before the work begins.

Why Broad Interest Doesn't Translate Into a Research Question

An interest is a direction. A research question is a destination. Confusing the two is the single most common reason proposals get sent back for revision.

Take "misinformation" as an example. It's a legitimate and urgent area of concern, but on its own it isn't researchable it's a field, not a question. Nobody can study misinformation in general any more than a biologist could study "life." What can be studied is something narrower sitting inside that field: how a specific platform's design features influence how quickly false claims spread during a particular type of event, or how fact-checking labels change reader behaviour when the reader already distrusts the source doing the labelling.

The instinct to stay broad usually comes from fear of narrowing too early and missing something important. In practice, the opposite happens. Broad topics produce shallow, repetitive work because there's nowhere for genuine analysis to take root. Narrow, well-defined questions are what allow a researcher to say something that hasn't already been said a hundred times.

Finding the Question Underneath the Topic

A useful way to move from topic to question is to notice friction the point where something doesn't add up, or where two things that should align don't. Journalism as a field is full of this kind of friction right now. Reporting practices built for print and broadcast are being applied, often badly, to platforms that behave nothing like either. Surveys keep suggesting that audience trust is sliding even as fact-checking infrastructure expands, which is itself a strange enough pairing to sit with for a while. Editorial judgement is increasingly shaped by algorithms nobody in the newsroom fully controls.

Each of these frictions contains a question, but the question only becomes visible once you ask what specifically is being contradicted, and for whom. Is trust falling among all audiences, or specifically among younger ones who encounter news through social feeds rather than mastheads? Is algorithmic influence changing what gets covered, or only how it gets framed once a story is already assigned? These distinctions matter enormously, because a question aimed at the wrong layer of the problem will produce findings that are technically correct but practically uninteresting.

It also helps to read not for content but for gaps. When you come across a study, an industry report, or even a well-argued opinion piece, ask what it assumes rather than what it argues. Almost every piece of writing on digital journalism assumes something about audience behaviour, technological neutrality, or newsroom incentives that the author never actually tests. That assumption is often a better starting point for a research question than the article's stated conclusion.

Testing Whether an Idea Is Actually Researchable

Once an idea starts to take shape, it needs to survive a few honest tests before it's worth committing to.

The first is whether it can be investigated with the access and methods realistically available. A fascinating question about internal editorial decision-making at a major broadcaster is worthless if no one there will speak to you and no documents exist to analyse instead. Ambition has to be matched against what can actually be observed, interviewed, coded, or measured within the time and access a student or academic realistically has.

The second test is whether the answer is genuinely unknown, rather than merely unstated. There's a difference between a question nobody has formally studied and a question everybody already knows the answer to but hasn't written a paper confirming. Asking whether tabloid headlines use more emotionally charged language than broadsheet ones, for instance, might be technically unanswered in a specific dataset, but most people who've read both would guess the direction of the answer before you start so the research would need to add a genuinely new angle, such as how that gap has changed over a specific period or in relation to a specific platform, to justify the effort.

The third test, and the one people skip most often, is scale. A question that sounds precise "how misinformation spreads on social media" is still enormous once you try to design a study around it. What platform. What kind of misinformation. What time period. What population of users. A workable question answers most of these before the research starts, not during it.

Where Weak Ideas Usually Go Wrong

Most weak proposals share a similar flaw: they describe a phenomenon rather than asking a question about it. "Fake news and its impact on democracy" describes a topic area. It doesn't tell a reader, or the researcher themselves, what specifically will be measured, compared, or explained. The fix is almost always to add a mechanism impact through what pathway, on which group, measured how.

Another common failure is choosing a question shaped by ideology rather than curiosity. It's tempting to set out to prove that a particular platform, industry practice, or political actor is responsible for a problem that already troubles you. Research designed to confirm a conclusion tends to produce weak, defensive work, because every piece of contrary evidence starts to feel like a threat rather than useful data. Stronger work comes from questions you're genuinely uncertain about, where the answer could reasonably go either way.

There's a third issue, and it's the one nobody warns you about until you're already three weeks into a project you no longer like. It's easy to pick a direction because it sounds impressive in a supervisor meeting, rather than because you can actually stand to live with it for months. Anyone trawling through journalism research topics looking for a dissertation subject will notice how many of them read well as a sentence but say nothing about what the daily work will feel like and that gap is exactly where projects quietly die. A subject that seemed exciting in a seminar can feel unbearable by the third week of coding interview transcripts. What actually carries a project through that tedious middle stretch isn't how good the idea sounded on day one; it's whether you're still curious about the mechanics of the problem once the novelty has worn off.

Narrowing a Broad Interest Without Losing What Made It Interesting

The instinct many students resist is cutting away the parts of an idea that first excited them, worried that narrowing will make the work feel smaller or less meaningful. Usually the reverse happens. A question about "how journalists use social media" becomes far more interesting once narrowed to how local crime reporters verify eyewitness footage shared on social platforms during breaking news events because now there's a specific decision-making process to trace, specific people to interview, and a specific set of pressures to examine.

The trick is to keep the underlying concern intact while shrinking the frame around it. If what genuinely interests you is the erosion of trust between audiences and institutions, that concern can live inside a very small, specific study one regional newspaper, one controversy, one six-month window without losing its connection to the larger question that made you care in the first place. The broader significance doesn't need to be baked into the question itself; it belongs in how you explain why the question matters once you've answered it.

What Changes Once You Approach It This Way

Treating topic selection as a process of building a question, rather than picking one off a shelf, changes the entire trajectory of a project. It means spending real time reading critically before committing to a direction, being willing to abandon an idea that doesn't survive the researchability tests, and accepting that the most promising questions are often narrower and less dramatic-sounding than the ones that first come to mind.

It also means the finished research tends to hold up better under scrutiny, because it was built to answer something specific rather than to illustrate something already believed. Reporting practices, misinformation, digital journalism, and audience behaviour will keep producing new friction points for as long as media systems keep changing faster than the institutions studying them. The researchers who do useful work in that space aren't the ones who found the most dramatic topic. They're the ones who learned to ask a question small enough to actually answer.