No articles match
Getting started with respondeR1 months ago
The idea | Data format | A first analysis | Which way is "better"? | Per-study results and a forest plot | Random effects and heterogeneity | How much does the MID matter? | Summary-of-findings rows | A threshold-free alternative | A real example: VAS pain after exercise therapy | Fitted meta-analyses | The Shiny application | Where next
Methodology1 months ago
The cut-point approach | Sampling variance of a responder probability | The pooling methods | Individual (recommended) | Weighted mean | Unweighted mean and median | Baseline risk: matched or median control | Relative effect measures and the NNT | Summary-of-findings rows | Common-language effect size (threshold-free) | The SMD bridge (method = "smd") | Converting a pooled effect fitted elsewhere | Random effects and heterogeneity | Uncertainty in the MID | Multi-arm trials | Final values and absolute thresholds | Interval options | Assumptions and limitations | Choosing a method | Changes from version 0.1.0 | Relation to the thesis | References
Responder analysis from a fitted meta-analysis1 months ago
From a meta (metacont) object | From a metafor rma object | From a bare estimate | How the conversion works, and what it assumes | Coming next | References
ASAP and STAR Methods profiles2 months ago
Available profiles | Project to the ASAP layout | Lossy fields | Licensing and attribution | Round-trip through the ASAP format
Extending krt with plugins2 months ago
A custom validator | A custom institutional profile | A custom resolver or suggest source
Extracting a KRT from a manuscript2 months ago
Scan text for identifiers | Extract a table | Read structured documents | LLM extraction (optional)
Getting started with krt2 months ago
Build a table | Validate | Normalize and export | Render for a manuscript | Provenance
Getting started with retraction4 months ago
Checking a file | Checking identifiers or a data frame | PubMed Central articles | Choosing sources | Offline mode | Identifier normalization
How the matching thresholds were calibrated4 months ago
The labeled corpus | Result 1: the assertion gate never false-accuses | Result 2: 0.90 is the empirical sweet spot for surfacing | Conclusion | Reproducing
Workflows: CI gates, reviews, monitoring, and pipelines4 months ago
Gate a manuscript render | Fail CI on retracted citations | Systematic reviews | Monitor a bibliography over time | Multiple sources | Interpret and export | Offline and at scale | Reproducible pipelines with targets