What do blind peer review, conflict of interest and rebuttals mean in practice?
Four terms that appear in every vendor’s feature list and every program committee’s requirements, and that mean different things to different people. Each is defined here by what has to happen inside the software for the claim to be true, with the questions that expose the difference. This page names no vendor; Rostrum appears only as a worked example, with a link to the manual page that documents the behavior.
1. What does “blind peer review” hide, and from whom?
Single-blind review hides the reviewers from the authors. Double-blind review also hides the authors from the reviewers. Scientific societies almost always mean the second when they say “blind”, because the integrity claim is that a reviewer scored the science and not the name at the top of it.
In practice “hidden from reviewers” has to hold in five places at once: the submission form, the abstract text, any uploaded file and its filename, every email the reviewer receives about the assignment, and every export or report a committee member can open. A system that masks the form and leaks the filename is not blind.
It also has to be enforced by the server rather than merely hidden by the interface. If the author’s name is present in the data sent to the reviewer’s browser and only concealed by the page, a curious reviewer is one keystroke away from it.
In practice: Rostrum’s blind review setting removes author names, affiliations, the submitter’s identity and any disclosures from everything a reviewer sees — on screen and in the assignment email — and a separate setting decides whether reviewers see supporting files at all, because a figure can carry a name the abstract text does not.
2. What is a conflict of interest, and what should the software do about one?
There are two halves, and they are often confused. Disclosure collection asks submitters and speakers to declare relationships — funding, employment, financial interests — so the program can be accredited and the audience informed. Reviewer conflict handling stops a reviewer scoring work they have a stake in: their own, a co-author’s, a colleague’s, a competitor’s.
The first half is a form and a report. The second half is a rule inside assignment, and it has two triggers: the system should never offer an abstract to its own authors, and a reviewer who opens an abstract and recognizes a conflict must be able to declare it, with a reason, and be permanently excluded from that abstract by every assignment mode thereafter.
A declared conflict should leave a trail. The committee needs to see who declared what and why, and the progress view needs to show the gap it left so it can be filled.
In practice: Rostrum removes the submitter and every listed author from the candidate list before any assignment is made, so self-review is impossible by construction rather than by policy; a reviewer who declares a conflict is excluded from that abstract for good, the declaration and its reason are kept, and the gap shows up in the progress view.
3. What is a rebuttal, and how is it different from a revision round?
A rebuttal is the author’s reply to the reviews before the decision is made: the reviews go to the author, the author answers point by point, and the committee decides with both in front of it. It is common in computing conferences and rare in medical and scientific societies, where the volume — thousands of abstracts, a few weeks — makes an author round before every decision impractical.
A revision round is the society-shaped alternative. The committee sends an abstract back once, asking for specific changes rather than accepting or rejecting it; the author revises within a deadline; the reviewers see the resubmission with the changes marked and their earlier answers to hand; the next decision is final.
Whichever a program uses, two things matter: the abstract must be locked while reviewers are scoring it, so every reviewer reads the same text, and the author must never be able to see a decision the committee has not yet chosen to send.
In practice: Rostrum runs the revision-round model: one revision request per abstract, a deadline you set, the choice of the same reviewers with their earlier answers kept or fresh ones, and decisions that stay staged and invisible to submitters until the committee sends them. A formal rebuttal round is not a step in that flow.
4. What is the difference between a comment to the author and a committee note?
Reviewers write two kinds of text. A comment to the author is meant to travel: the committee can forward it with the decision, so the author learns why. A committee note is meant to stay: it is the reviewer speaking to the program committee and must never leave staff under any circumstances.
The distinction is only real if the system keeps the two apart at the data level and the decision letter can only ever include the first kind. Ask to see where each is entered and where each can appear.
In practice: In Rostrum a reviewer fills in the scoring questionnaire, gives a recommendation, and may write both a comment to the author and a committee note; the first can be forwarded with the decision, the second cannot leave staff.
What to ask any vendor
- Show me, on a reviewer’s screen and in a reviewer’s email, everything the reviewer can see about an author. Then show me the export.
- Is author identity removed on the server, or hidden by the page?
- What happens when a reviewer opens an abstract by a colleague? Can they declare it, and does the system remember?
- Can an author ever see a decision before the committee sends it?
- Where does a reviewer write something the author must never read, and how is it kept out of the decision letter?