Conference literature reviews don't need to be exhaustive —they need positioning

Start with the premise.

At a journal club, your job is to introduce someone else's paper. At a conference, the star is your own research or case. Prior research isn't the protagonist — it's a supporting cast that exists to explain why your presentation needed to happen.

Miss this distinction and approach the literature the way you would for a journal club, and your review of prior work runs long, eating into the time for your own findings. Supporting material should be brief.

In the background of a conference talk, there's ultimately only one thing the audience and the session chair want to know.

Why was this research necessary?

Answering that doesn't require an exhaustive list of the literature. It requires clear positioning: "This much is known. This much is not. That's why we did this study."

This isn't a contest for the most complete literature review, so a small, carefully chosen set of citations is perfectly fine. How the dots connect matters far more than how many there are.

So how do you organize the papers you've collected? My recommendation: sort them into three boxes.

Box 1: similar studies — work that tackled the same question as yours. Check whether their results agree with each other or diverge.

Box 2: different studies — a similar question, but different populations or methods. Differences like "never studied in the elderly" or "no data from this region" become the direct justification for your own study.

Box 3: the blank space — what you searched for and couldn't find. The very fact that nothing turned up becomes a powerful line in your presentation.

The sorting criterion is not "is this a good paper?" but "how far is it from my own research?" This is a way of reading specific to conference preparation — distinct from both critical appraisal and a general literature review.

With the three boxes filled, writing the background becomes far easier.

Line one: what we know so far. Line two: what we don't yet know well enough. Line three: so here's what we investigated.

Known, unknown, therefore this study. With those three lines as the skeleton, your background holds together whether it's spoken, on slides, or in the abstract.

Conversely, any literature that doesn't fit those three lines — however interesting — takes courage to leave out of the presentation. Leave it out.

There is one step here you absolutely cannot skip.

For the references that actually appear in your background — the handful you cite — verify the numerical results and the scope of the conclusions in the original PDFs, not in AI summaries.

In conference Q&A, questions sometimes target your prior-research claims: "About that study you just cited..." If you've been citing secondhand or leaning entirely on AI summaries, this is where you get stuck.

The ethics of a conference literature review, I'd argue, is choosing only a few citations and taking full responsibility for those few.

That said, you don't need to read all 20 collected papers closely.

At the sorting stage, grasping each paper's aim, population, and main result from an AI summary is enough. Sort using summaries; once you've decided the few papers you'll cite, switch to a close reading of those originals.

Sorting is the AI summary's job; citing is done by hand. That division of labor is what's realistic on a deadline.

And if a summary leaves you thinking "this paper's positioning is fuzzy," jot that unease into the summary as a note — it will pay off later.

In Paperfy, adding your collected PDFs to the library gives each one a paper page with an AI summary and extracted figures.

During sorting, you can add positioning notes directly to each paper's summary — "Box 1: differs from ours in target age group" — while reading. Because AI summaries are editable by hand, recording each paper's distance from your own research uses the same workflow as correcting a misreading.

Before the conference, you can turn the key points into a radio script and review it while commuting, with your three background lines in mind. The script is editable too, so you can reshape it to match the flow of your background explanation and use it to rehearse the spoken version.

The goal of organizing the literature isn't to produce an excellent review — it's to find the spot where your own research stands out.

Sort the collected papers into three boxes: similar, different, and blank. Tell the background in three lines: known, unknown, therefore this study. Verify every cited paper in the original PDF.

Start by sorting the PDFs you've already collected into the three boxes.