Bayram Eker
Menu
EN
CV Start a conversation

Home / Writing / Research

What I look for when reading a preprint

A preprint lets us examine work before peer review. Knowing what its links and identifiers mean helps us read both the contribution and its limits.

Evidence review interface showing provenance and verification state
Following a result back to its source is part of reading it. QFlow evidence-review interface.

When I open a research paper, I want to know how the author reached the result. Before I get that far, the page gives me a different kind of information: an arXiv number, a DOI, perhaps a repository. Those links help me find and return to the work. Understanding what they do, and what they cannot tell me, is part of reading it fairly.

At the time of writing, my own publications this year are preprints or a self-published book: two arXiv papers, an SSRN architecture paper with a Zenodo deposit, and a monograph through my own press. None has been peer-reviewed. I want that to be easy to see, so readers know how to approach the work.

arXiv makes a manuscript available with a timestamp and identifier. A DOI gives it a persistent address. Zenodo provides an archival deposit. SSRN distributes and indexes work in its subject areas. These are useful services for finding and citing material; they are different from an expert assessment of its conclusions.

The same care applies to a book listed on Amazon or a profile on ResearchGate. Availability is valuable, but it should not be mistaken for endorsement. On my own site, I try to describe what each service contributes rather than let its name carry a larger claim.

Read it as work you can question

A preprint is still something we can learn from. The author has made a method and argument public, and others can examine them. I start with the question, the comparison and the conditions. Then I look for what the data supports and where the author is making an interpretation.

For my own work, publishing early gives others a chance to respond while the questions are still developing. The trade-off is less external validation. That makes a clear account of demonstrated results, plausible extensions and speculation especially important.

The question I bring back to my own drafts

What would a careful reader need before trusting this conclusion? Sometimes the answer is another experiment. Sometimes it is simply a sentence explaining a condition I had taken for granted. Writing that sentence is part of the research, because it gives another person a fair chance to judge it.

I do not expect every preprint to be a finished story. I do expect enough detail to understand what has been done and what remains open. That is the standard I want readers to be able to apply to mine.

Three questions for a first reading
  • What is the question?
  • What supports the answer?
  • What is still open?

Written by Bayram Yüksel Eker. Corrections and disagreement are welcome at bayram@neuraparse.com.