Independent edition · GNK ASG network
SCENA ZNANOSTI — how science actually happens
Science is not a series of discoveries but a series of checks. This desk follows the process: who reviewed a paper, whether it was published before review, what was corrected afterwards, and why something was retracted. It is less exciting than headlines — and far more useful.
The desk does not run research or peer-review papers. It carries published documents and marks what is peer-reviewed and what is not.
What this edition follows
What a reviewer actually checks — and what they don’t
A reviewer does not repeat the experiment. They read the method, check the logic and ask whether the conclusion is larger than the data. That helps…
Preprints: speed that requires a thinking reader
A preprint is a public paper before review. It is not less valuable — only less verified. The difference is that the reader must carry part of the…
Reproduction: the most important experiment is someone else’s
A result nobody can repeat is not a result but an event. Reproduction is therefore the least attractive and most important part of science.
How SCENA ZNANOSTI is made
One rule: the genre is never hidden. A preprint is a preprint, a reviewed paper is a reviewed paper, and a press release is a press release.
01A headline is not a conclusion
A study of 12 people and a study of 12,000 can end with the same headline. The difference is that the first cannot be repeated, and the second can — which is why “how many people” is the first question of any serious conversation.
So alongside every topic this desk publishes three things: what was measured, on how many participants, and where the paper appeared. If any of those is missing, the topic is not ready.
- Peer review: what a reviewer actually checks
- Preprints: speed without verification, and why that helps
- Reproduction: the most important experiment repeats someone else’s
- Corrections: what it looks like when science fixes itself
02Two speeds, one goal
Preprint servers sped up the sharing of results: a paper can be public today and reviewed in a few months. That helps in a crisis and in collaboration, but it also obliges the reader to check whether review happened.
So on this site a preprint is never labelled a “study” without saying so. The difference between “published” and “peer reviewed” is not a formality but information.
Terms, without fog
Ten terms that decide how a story reads. Each definition is short on purpose.
Peer review
Independent experts read a paper before publication and recommend changes or rejection.
Preprint
A paper posted before review, usually on a server such as arXiv; it is not proof of verification.
DOI
A persistent identifier that leads to the paper or its record, independent of URL changes.
Retraction
An editorial statement that a paper no longer stands; the reason may be error, misconduct or non-reproducibility.
Correction
A fix to part of a paper that does not destroy the conclusion but changes a figure or an image.
Reproduction
Repeating an experiment from the paper’s description; success means the result is not accidental.
Sample size
The number of units in a study; a small sample carries large uncertainty, whatever the result.
Conflict of interest
A financial or personal link between authors and the outcome; it must be declared.
Blinding
A procedure where participants or researchers do not know who is in which group.
Preregistration
A publicly recorded hypothesis and analysis plan before data collection.
Questions that keep coming back
Does peer review mean a paper is correct?
No. Review is a pre-publication check, not a guarantee. Retracted papers often passed review — which is why what happens after publication matters.
Are preprints bad?
No. A preprint is fast and open, but unverified. It is useful as a source of questions and an early look, not as final proof.
Sources
Every source below is public and checkable. Where a page cites a figure, the figure comes from one of these documents — and the document has a date.
- arXivThe largest preprint server; papers before review, clearly marked as not peer reviewed.
- NatureJournal and an explanation of its own editorial process.
- ScienceJournal and its corrections and retraction policy.
- Retraction WatchA database of retracted papers and the reasons behind retractions.
- PubPeerPublic post-publication discussion of published papers.
- DOI FoundationA persistent identifier; a link that does not depend on the journal’s URL.
- Hrvatska zaklada za znanostNational research funding and public project information.
Other editions in the network
Each edition has its own domain, its own archive and its own editor’s signature — and all of them point back to the same source of record.
BOR USA
Combat sports in the United States: regulation, safety and how to read a fight.
KOŠARKA 360
Basketball without noise: how to read a season, what the numbers count, and why the regional league matters.
STEP BUDUĆNOSTI
Dance, step and movement as technology: sport, injury, motion capture and the stage.
ZDRAVLJE 360
Health without miracles: prevention, sleep, movement, and how to read a medical study.
IZBORI U SRBIJI
A monitoring desk: what media in Serbia, Croatia, Bosnia and the US publish about the election — with source and date.