Table 2.
Variable | Point (CI: 95%) and range by method |
|||||||
---|---|---|---|---|---|---|---|---|
Logistic | Poisson | Robust Poisson | Log-binomialb | Bayesian log-binomial (MCMC)a | Δlgt%c | Δpoi%d | Δbln%e | |
Routine use of rectal examination | 4.35 (2.52; 7.49) 4.97 | 2.82 (1.81; 4.39) 2.58 | 2.82 (1.96; 4.06) 2.09 | 2.56 (1.78; 3.67) 1.88 | 2.47 (1.78; 3.72) 1.94 | 69.91 | 10.15 | −3.51 |
Direct contact over the fences | 2.04 (1.22; 3.41) 2.19 | 1.64 (1.06; 2.54) 1.47 | 1.64 (1.13; 2.39) 1.26 | 1.62 (1.12; 2.34) 1.22 | 1.51 (1.14; 2.41) 1.27 | 25.93 | 1.23 | −6.80 |
Farms that do not use artificial insemination (AI) | 3.01 (1.60; 5.66) 4.06 | 2.07 (1.28; 3.34) 2.06 | 2.07 (1.41; 3.03) 1.62 | 1.78 (1.27; 2.49) 1.21 | 1.68 (1.21; 2.38) 1.17 | 69.10 | 16.30 | −5.61 |
aCredibility interval: 95%.
bReference model.
cRelative difference in point estimates between the logistic and reference model.
dRelative difference in point estimates between the Poisson, robust Poisson, and Reference Model.
eRelative difference in point estimates between the Bayesian log-binomial, and reference model.