Feedback and rating guide

Independent ratings, privacy choices, aggregation, and your contribution export

Back to dashboard

Why rate independently?

The standard dashboard is useful for discussion, but its existing scores and analyses can anchor a new judgment. The independent-rating view hides priority scores, recommendations, human feedback, and AI prioritization analysis until you submit your own rating for that paper in this browser.

Papers are always ordered by date, newest first, in the independent view. Abstracts and source metadata remain visible so you can make an informed first pass.

Open the independent-rating view

Quick ratings and percentile equivalents

The five-point scale answers: “How strongly should The Unjournal commission an independent evaluation?” For aggregation and sorting, these choices are translated to percentile-equivalent scores:

-- = 10Strong No
- = 30Weak No
~ = 50Neutral
+ = 70Weak Yes
++ = 90Strong Yes

Detailed ratings can use any 0–100 percentile. A later rating from the same browser revises that paper’s current rating rather than intentionally adding another current vote.

What is published and what remains private?

ContributionDefault handling
Numeric ratingIncluded only in privacy-safe public aggregates; the public dashboard does not publish the rater’s identity beside a paper.
DiscussionYou choose public anonymous, public with a chosen name or pseudonym, team-only, or confidential administration-only visibility.
NamePublished only when you explicitly choose named public discussion and provide the display name.
Email, browser key, submission IDNever published on the dashboard.

How the human–AI synthesis works

Each person’s latest rating for a paper counts once. Quick ratings have weight 1. A full rating has weight 2 without detailed discussion, weight 3 for a public rating with detailed discussion, and weight 4 for a team rating with detailed discussion. “Detailed” currently means at least 120 characters or 20 words.

The AI score has weight 3 in the same quick-rating equivalents. Papers without human feedback retain their AI score. These weights are intentionally described as provisional and should be revisited as evidence accumulates.

Download your contributions

You can download the ratings, category judgments, and discussion text retained in this browser as JSON. The export does not send anything to The Unjournal or another service.

Request a copy of server-held submissions

Browser-local export cannot include submissions made on another device or records no longer retained by this browser. For a broader server-held history, contact The Unjournal using the same email address you supplied with your contributions.