LTRC conference corpus
Description
This dataset catalogues and thematically codes every available conference programme of the Language Testing Research Colloquium (LTRC), the field's principal annual research conference, from its first meeting in 1979 through 2026. It was compiled to support empirical claims about how the field of language testing has changed since the adoption of the first International Language Testing Association (ILTA) Code of Ethics in 2001.
Each of the 3,506 rows represents one individually titled contribution (paper, poster, work-in-progress, symposium contribution, pre-conference workshop, or keynote/plenary address) drawn from 46 of the 47 LTRC editions held to date (the 2004 programme could not be recovered; the 2002 programme survives only in part). Every contribution is assigned a primary research theme from a fixed taxonomy of 27 categories developed inductively from the corpus itself (e.g., Speaking assessment, Automated scoring/NLP/AI, Validity & Validation, Testing policy & political use), with a secondary theme assigned where a title clearly spans two categories.
Thematic coding was performed by large language model (Claude, Anthropic) agents operating on presentation titles and immediate programme context. Coding reliability was assessed against a blind, independent human recoding of a random 200-item sample (82.5% observed agreement; Cohen's κ = .82); the human rater's classifications are included alongside the original coding for full transparency and reanalysis.
Contents:
- The anonymized coded corpus
- R script reproducing all statistics and figures reported in the manuscript
- Appendix with 27-category taxonomy with definitions and frequency counts
- Readme-file with full column dictionary, coding methodology, and coverage limitations
Known limitations: the 2004 LTRC programme is entirely absent (not recoverable from any known source), and the 2002 programme is only partially represented. All statistical comparisons in the analysis script account for this explicitly, with a sensitivity check excluding the partial 2002 programme reported for headline findings.
Files
appendix_taxonomy.csv
Files
(1.0 MB)
| Name | Size | |
|---|---|---|
|
md5:a05c827bb029b77d7518ed381a6e8ace
|
5.4 kB | Preview Download |
|
md5:f6010313268270ba49fcde7d69201ede
|
1.5 kB | Download |
|
md5:3f42c70db1f1ae256f294c68f91fd111
|
943.3 kB | Preview Download |
|
md5:93b10463720cadb8b228c371978836ae
|
52.6 kB | Download |
|
md5:2d9f341a3242fb02cbce1494c01e0ded
|
10.0 kB | Preview Download |
Additional details
Dates
- Created
-
2026-08-26
Software
- Programming language
- R