Radar Perene / Archive / science

Zotero for the researcher who studies markets alone

◦ Index methodology v2.2 (working papers with DOI). See the methodology.

Science

Every research effort that lasts longer than a month faces the same silent adversary: the researcher's own memory. The PDF downloaded in March cannot be found again in September. The claim that seemed unforgettable — "I read that in a paper on banking spreads" — loses its origin, and a claim without an origin cannot be used. Inside a university, the surrounding infrastructure cushions the problem: there is a library, a repository, colleagues who remember. Whoever researches markets on their own has none of that. They have, at most, a folder called "papers" with two hundred cryptically named files.

Zotero is a free, open-source reference manager: a personal catalog in which every study, book or document enters with its full metadata — authors, year, journal, DOI — and out of which citations come formatted, verifiable and findable years later.

The description is modest on purpose. Zotero does not think, does not summarize, does not read for anyone. It does a single thing, and that thing is the difference between an archive and a pile: it turns "I read it somewhere" into "it is here, with date, author and identifier".

The specific problem of research outside the university

A researcher with an institutional post cites poorly and the structure corrects them: the advisor demands, the referee flags, the library helps. An independent researcher cites poorly and nobody notices — until the day the work is read by an outsider, at which point the loose citation costs exactly what the independent researcher has least of, which is credibility.

That is why the discipline of references weighs more, not less, on whoever works alone. An independent study of the Brazilian market will be read with initial suspicion; a traceable bibliography is one of the few proofs of seriousness a reader can check in minutes. Every citation with a DOI is an invitation to verify — and research that invites verification belongs to another category of text.

How the house library is organized

Radar Perene keeps a single master library in Zotero, and maintaining it has taught us more about method than we expected. The structure can be described without secrets.

First: one library only, with isolated subtrees per line of work. The market-research line lives in its own subtree, marked with a house tag — which lets other fronts share the same vault without contaminating searches. The alternative, one library per project, was discarded because the finance literature repeats across projects, and duplicating entries is manufacturing inconsistency.

Second: the rule of "one work, one entry". The same study found twice — once as a working paper, once as a published article — becomes one canonical entry, with a note pointing to the earlier version. Without that rule, the catalog bloats and citations drift apart.

Third: the DOI is the key. Every entry that has a persistent identifier comes in with it; those that arrived without one — old captures, institutional documents, reports — are flagged as pending, and completing that metadata is declared maintenance routine, not hidden embarrassment. An honest catalog knows what it lacks.

Fourth: Zotero holds other people's literature. The house's own output lives elsewhere — in a public repository with DOIs, the subject of the next article on this trail — and the separation is deliberate: a reading catalog and a record of production are different functions, with different demands of permanence.

What the tool does not solve

No reference manager decides what deserves to come in. The collector's temptation — save everything, read later — merely changes shape: the chaotic folder of PDFs becomes a chaotic catalog of PDFs, prettier and equally useless. The house library grows slowly on purpose: what enters is what has been read or what sits in the short reading queue, and the short queue has a finite size.

Nor does the tool replace judgment about the quality of sources. An impeccable catalog of weak studies produces weak research impeccably cited. The criteria the house applies before letting a source in — who published it, on what data, with what conflicts — are the subject of another trail in this section, on reading market science without being fooled.

Frequently asked questions

Is Zotero really free?

The program is free and open source, maintained by a nonprofit organization. What can be paid for, optionally, is attachment-sync storage beyond the free quota — the metadata syncs at no cost.

Does it work without a university affiliation?

Entirely. No central feature requires an institutional e-mail or a database subscription. What the affiliation grants is access to paywalled articles — a distinct problem, which open repositories and working papers soften.

How is it different from an ordinary file manager?

Metadata and citation. A folder stores the PDF; Zotero stores the PDF tied to author, year, journal and DOI, and returns that as a formatted citation in any style. The folder answers "where is the file?"; the catalog answers "what exactly am I citing?".

Is it worth it for someone who never intends to publish?

The function is the same for someone who only reads: finding again, comparing and checking. Anyone who follows market research with any seriousness accumulates sources the same way — and loses sources the same way.

---

The catalog organizes what the house reads. The next question on this trail is where what the house writes lives — and why the choice of repository, and even of license, is also a research decision: What is Zenodo.

House reading: what this discipline of sources produces every day is in today's reading, in the Diário.

Building a research library for a specific question is bench work — the kind the house details in conversation, not in an article.