Share of articles mentioning slaves or public/private bondspeople, by reign, with sample precision.
1.8%of articles have a slaves keyword · precision 0.68
Key point
Re-coded in 1.2 with a narrower codebook, the slaves keyword has 68% precision (blind re-coding κ 0.86). About a third of keyword articles are not about slaves, so multiply counts by the precision.
Formula
Keyword articles / all articles
How to read it
How much of the state record concerned the status system over time.
Caveats
False hits such as '奴' inside names were removed by sample coding. Under the 1.0 codebook (v1) an independent coder re-coding 30 articles agreed on 22 (κ 0.44): the scope of 'slaves as a topic' was read differently. In 1.2 the 102 sampled articles were re-coded with a v2 rule naming four contexts as yes (punishment, interrogation, transfer, labor duty): κ 0.86 on 30 blind articles, precision 0.54 → 0.68.
Reliability card
sampledSampled precision — share of keyword hits that mean what the metric says, measured on a coded stratified sample
Precision 67.6% (Wilson 95% 58.1%–75.9%) on a coded stratified sample of 102 articles (10 per 50-year period, principal editions). The 8 articles left undecided were counted as 'no'. The codebook and all judgments can be downloaded from the Data page.