Mahmood Ahmad
Tahir Heart Institute
author@example.com

Protocol: Benford Screening of 1.2 Million Meta-Analytic Values Finds No Corpus-Level Digit Anomaly

This protocol describes the evidence synthesis for Benford Screening of 1.2 Million Meta-Analytic Values Finds, targeting reproducible estimation of Mean absolute deviation (MAD) in a versioned workflow. Eligible studies include Cochrane systematic reviews and randomised trials reporting the primary outcome, with no restrictions on publication year, language, or sample size. Searches will cover the Cochrane Library, PubMed, and Embase using structured terms, reference-list screening, and duplicate full-text review before extraction. The primary analysis will estimate Mean absolute deviation (MAD) using restricted maximum likelihood random-effects meta-analysis, reporting 95 percent confidence intervals, prediction intervals, and prespecified model checks. Heterogeneity will be summarised using I-squared and tau-squared, with sensitivity analyses across variance estimators, exclusion scenarios, and leave-one-out patterns. Analysis code will be versioned and archived at https://github.com/mahmood726-cyber/benfordma, and reporting will follow PRISMA 2020 guidance to support independent verification and reuse. Anticipated limitations include publication bias, clinical heterogeneity, sparse data in some settings, and the constraints of aggregate-level evidence synthesis.

Outside Notes

Type: protocol
Primary estimand: Mean absolute deviation (MAD)
App: BenfordMA v1.0
Code: https://github.com/mahmood726-cyber/benfordma
Date: 2026-03-26
Validation: DRAFT

References

1. Carlisle JB. Data fabrication and other reasons for non-random sampling in 5087 randomised, controlled trials in anaesthetic and general medical journals. Anaesthesia. 2017;72(8):944-952.
2. Brown NJL, Heathers JAJ. The GRIM test: a simple technique detects numerous anomalies in the reporting of results in psychology. Soc Psychol Personal Sci. 2017;8(4):363-369.
3. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Introduction to Meta-Analysis. 2nd ed. Wiley; 2021.

AI Disclosure

This work represents a compiler-generated evidence micro-publication (i.e., a structured, pipeline-based synthesis output). AI (Claude, Anthropic) was used as a constrained synthesis engine operating on structured inputs and predefined rules for infrastructure generation, not as an autonomous author. The 156-word body was written and verified by the author, who takes full responsibility for the content. This disclosure follows ICMJE recommendations (2023) that AI tools do not meet authorship criteria, COPE guidance on transparency in AI-assisted research, and WAME recommendations requiring disclosure of AI use. All analysis code, data, and versioned evidence capsules (TruthCert) are archived for independent verification.