Primary-study data for meta-analytic SEM, extracted and verified.
MASEMiner reads your primary studies and returns correlations, correlation matrices, factor loadings and study metadata as structured data. Every value is shown next to the highlighted passage it came from, so you can check and correct it before you export.
You'll need
- PDFs of the primary studies (one batch at a time)
- Your own OpenAI, Google, Anthropic, Mistral or DeepSeek API key, a self-hosted model, or MASEMiner credits with an account
- A minute or two per paper while the model reads it
What you'll get back
- Correlations between the constructs you name, per sample, with reliabilities
- Item-level factor loadings and inter-factor correlations for indirect information
- Study metadata: N, age, % female, country, instrument, analysis method
- Verbatim evidence with page references, and a confidence rating per block
- JSON and CSV export including your corrections, with the preset and model recorded
How it works
- Name your effect sizes and variables once; save the setup and reuse it across batches
- Review each paper against the PDF, verify or flag each row, edit values in place
- Papers imported without a run keep their citations and can be reviewed the same way
MASEMiner is a product surface of the open-source Metalens extraction engine.
If you prefer to keep PDFs on your local machine, you can run MASEMiner locally.