The data | Step 1: write a codebook | Step 2: code the text | Step 3: check the coding against a human | Step 4: link themes to the metrics | Step 5: plot | Segment-level transcripts | Data protection
Package: saydo 0.1.0
Bernard Isekah Osangir
saydo: Link What People Say to What They Do with LLM-Coded Themes
Measures the say-do gap in mixed-methods data: whether themes in what people say go with differences in what they do. Codes free text against a codebook with large language models through 'ellmer', or any coder function, and validates the codes against a human-coded sample with Cohen's kappa, precision and recall. Links the codes to participants' behavioural or outcome data, aggregating segment-level transcripts to participants where needed, and tests each theme against each metric with Welch's t-test and multiple-comparison adjustment. Developed by Neudata Consulting Ltd.
Authors:
saydo_0.1.0.tar.gz
saydo_0.1.0.zip(r-4.7-any)saydo_0.1.0.zip(r-4.6-any)
saydo_0.1.0.tar.gz(r-4.7-any)saydo_0.1.0.tar.gz(r-4.6-any)
saydo_0.1.0.tgz(r-4.6-emscripten)
manual.pdf |manual.html✨
DESCRIPTION |NEWS
card.svg |card.png
saydo/json (API)
| # Install 'saydo' in R: |
| install.packages('saydo', repos = c('https://cran.r-universe.dev', 'https://cloud.r-project.org')) |
Bug tracker:https://github.com/neu-data/saydo/issues
Pkgdown/docs site:https://neu-data.github.io
Last updated from:f7631e480f. Checks:6 OK. Indexed: yes.
| Target | Result | Time | Files | Syslog |
|---|---|---|---|---|
| linux-devel | OK | 125 | ||
| source / vignettes | OK | 176 | ||
| linux-release | OK | 207 | ||
| windows-devel | OK | 72 | ||
| windows-release | OK | 82 | ||
| wasm-release | OK | 284 |
Exports:code_agreementcross_tab_themesllm_auto_codemock_codermock_mixed_dataplot_theme_impact
Dependencies:askpassclicorocpp11curlellmerfarverfastmapggplot2gluegtablehttr2isobandjsonlitelabelinglaterlifecyclemagrittropensslotelpillarpkgconfigpromisesR6RColorBrewerRcpprlangS7scalessystibbleutf8vctrsviridisLitewithr
Readme and manuals
Help Manual
| Help page | Topics |
|---|---|
| Measure agreement between LLM codes and human codes | code_agreement |
| Cross-tabulate qualitative themes against a quantitative metric | cross_tab_themes |
| Auto-code text with a codebook using structured LLM extraction | llm_auto_code |
| A keyword-based stand-in coder for demos and tests | mock_coder |
| Generate synthetic mixed-methods behavioural data | mock_mixed_data |
| Plot how a metric shifts when each qualitative theme is present | plot_theme_impact |