coreSR.AIcoreSR
Module

INSPECT-SR

Trial trustworthiness screening with a research-integrity toolkit

Screen the RCTs in a review for fabrication, falsification and catastrophic error with the INSPECT-SR checklist (21 items, 4 domains). Every trial is run through Retraction Watch and PubPeer, trial-registry retrieval with prospective-registration timing and registration-versus-publication comparison, statcheck-style p-value recomputation, GRIM/GRIMMER and SPRITE granularity tests, the Carlisle baseline p-value method, participant-flow reconciliation, loss-to-follow-up, allocation-split and recruitment-rate plausibility rules, ethics-reporting checks, tortured-phrase screening, a citation-integrity check that resolves the whole reference list against Crossref and PubMed to expose fabricated citations, and image forensics on the PDF. The AI proposes item responses grounded in that evidence; reviewers record the holistic domain and overall judgements. A trial × domain matrix, per-trial and project Word reports with the full audit trail, CSV export, a Researcher Allegiance Bias sibling, and a sensitivity re-run that drops flagged trials from a linked meta-analysis complete the module.

INSPECT-SR — Trustworthiness Dashboard

Key Features

INSPECT-SR instrument (21 items, 4 domains) with escalator handling
Fabricated-citation detector: every reference resolved against Crossref and PubMed
statcheck p-value recomputation, GRIM/GRIMMER, Carlisle baseline method, flow arithmetic
Retraction Watch, PubPeer, registry timing and registration-vs-publication diff
Tortured-phrase and image-forensics screening
Trial × domain matrix, Word and CSV reports with audit trail, sensitivity re-run engine

How It Works

1

Add trials

Bulk-upload PDFs or enter identifiers

2

Check

Lookups and integrity engines run automatically

3

Review

Confirm item, domain and overall judgements against the evidence

4

Report

Export reports or re-run the linked meta-analysis without flagged trials

Ready to get started?

Start your 7-day free trial and use INSPECT-SR today.

coreSR.AICoreSR

AI-powered evidence synthesis