The Unjournal  ·  joined-up evaluation hosting  ·  a rough experiment

A whole evaluation package on one page

When we evaluate a paper, the pieces — the paper, the evaluations, the ratings, the authors' reply — live on separate pages (individual "pubs" in our PubPub interface), and you have to jump between them. I've been trying out ways to put a whole evaluation package on one page instead. These are rough prototypes, meant to sit alongside PubPub rather than replace it. I'm not sure they're better — that's what I'd like your help with.

Most of the text here — the summaries and plain-language explanations — was AI-drafted and then reviewed and edited by me. The evaluations, ratings, and authors' responses are the evaluators' and authors' own words.

four real evaluation packages · the "hub" layout

Start here: four real evaluation packages, one page each — the "hub" layout

These are real, already-published Unjournal evaluations, each put onto a single page — the layout we call the "hub". All four include the authors' response, set out point by point against the evaluators' comments — which is the part that's hardest to follow when it's scattered across pages.

Empirical economics / IO Author response

Observational price variation does not reproduce experimental price elasticities

Bray, Sanders & Stamatopoulos (2026)

Two evaluators on whether the usual scanner-data methods recover the price elasticities you'd get from an actual pricing experiment. The authors reply point by point.

2 evaluators · Lars Roemheld + 1 anonymous

Open the Bray et al. hub →

Macro / economics of AI Author response

Artificial Intelligence and Economic Growth

Aghion, B. Jones & C. Jones (NBER, 2019)

Two evaluators (one of them caught a proof error) on a paper about AI, growth, and when you might get a "singularity". The authors reply.

2 evaluators · Benzell · Trammell

Open the Aghion et al. hub →

Conservation / policy Author response

Banning wildlife trade can boost demand for unregulated threatened species

Kubo et al.

Two evaluators on whether trade bans push demand onto related, still-legal species, using Japanese auction data. The authors answer each comment in turn — in unusual detail.

2 evaluators · Jia Huan Liew + 1 anonymous

Open the Kubo et al. hub →

Global health / RCT Author response

Intergenerational Child Mortality Impacts of Deworming (KLPS, Kenya)

Walker et al.

Two (anonymous) evaluators on whether deworming one generation lowers their children's mortality. One flags an apparent departure from the pre-registered plan; the authors explain why it isn't one.

2 evaluators (anonymous) · ratings on PubPub

Open the Walker et al. hub →

if you're curious

A few other formats I tried

The four pages above use the "hub" layout because, after trying a handful of approaches on one example (Benjamin et al. on wellbeing scale-use, evaluated by Kaiser and Prati), it seemed the most readable to me. The others each did one thing well, and you might prefer one of them — they're here if you want to compare. You can skip this part.

the one above used here

Reading Hub

Everything on one scrolling page, with the ratings up top and a link from each rating into the part of the write-up it refers to.

Open the Hub →

alternative

Claim-anchored reader

Two panes: pick a claim on the left, and see what each evaluator said about that claim on the right.

Open the claim-anchored reader →

alternative

Scrollytelling explainer

Walks you through the paper's idea first (with a couple of interactive bits), then the full record below it. Probably the gentlest if you're not a specialist.

Open the explainer →

alternative

Threaded dialogue

The whole package as one threaded conversation, labelled by who's speaking and filterable by topic.

Open the threaded view →