Skip to main content

Table 4 Estimates of unobserved contacts, population size and 95% CI (n = 341) for the three approaches (based upon the negative-binomial model)

From: Sensitivity of contact-tracing for COVID-19 in Thailand: a capture-recapture application

Estimators

\({\hat{f}}_0\)

\({\hat{N}}\)

Bootstrap median

95% CI normal approximation

CI from percentile BT

MLE

98.18

439.18

438.271

402.07–474.47

391.25–510.05

Turing

101.78

442.78

438.0309

401.35–474.72

390.74–511.09

Chao

148.84

489.84

439.5344

382.01–497.06

376.29–566.81

  1. Fifth and sixth column show bootstrap CI by the normal approximation and percentile methods, respectively