Table 3.
Final multivariable negative binomial regression model used to investigate the effect of different variables on the odds for a higher number of non-compliance during animal welfare control made by the CAB (County Administrative Board, official control) or Arla (on the private standard Arlagården) in the Swedish county of Västra Götaland during 2010–2013.
| Variable | CAB | Arla | ||||
|---|---|---|---|---|---|---|
| IRR | SE | p-Value | IRR | SE | p-Value | |
| Intercept | 1.56 | 0.30 | 0.019 | 0.77 | 0.11 | 0.071 |
| Season | 0.035 1 | <0.001 1 | ||||
| Winter (Dec–Feb) | 1 | 1 | ||||
| Spring (Mar–May) | 0.97 | 0.13 | 0.81 | 0.70 | 0.10 | 0.012 |
| Summer (Jun–Aug) | 0.46 | 0.12 | 0.004 | 0.44 | 0.10 | <0.001 |
| Autumn (Sep–Nov) | 0.88 | 0.15 | 0.47 | 0.88 | 0.17 | 0.38 |
| Control type | <0.001 1 | |||||
| Random | 1 | |||||
| Complaint | 1.33 | 0.26 | 0.14 | |||
| Risk-based | 1.12 | 0.19 | 0.50 | |||
| Thematic pasture | 0.44 | 0.15 | 0.02 | |||
| Thematic other | 0.90 | 0.14 | 0.52 | |||
| Cross-compliance | 0.26 | 0.11 | 0.002 | |||
| Others | 1.85 | 0.83 | 0.17 | |||
| Notification | 0.28 1 | |||||
| No | 1 | |||||
| Yes | 1.19 | 0.20 | 0.28 | |||
| Housing | 0.035 1 | 0.002 1 | ||||
| Cubicles | 1 | 1 | ||||
| Tie-stalls | 1.42 | 0.19 | 0.01 | 1.53 | 0.21 | 0.001 |
| Mixed | 1.15 | 0.40 | 0.69 | 0.39 | 0.31 | 0.23 |
| KRAV | <0.001 1 | 0.038 1 | ||||
| No | 1 | 1 | ||||
| Yes | 0.53 | 0.10 | 0.001 | 0.51 | 0.16 | 0.038 |
| Season × KRAV | 0.038 1 | |||||
| Winter or No | 1 | |||||
| Spring × Yes | 2.18 | 0.94 | 0.068 | |||
| Summer × Yes | 0.42 | 0.47 | 0.44 | |||
| Autumn × Yes | 0.60 | 0.31 | 0.31 | |||
1 Joint Chi-square test of effect.