Table 6 Significant predictors of perception accuracy and bias from binary (BLR) and multinomial (MLR) logistic regression models.
Dependent variable | Model | Outcome category | Predictor(s) | Odds ratio (95% CI) | p-value | Interpretation |
|---|---|---|---|---|---|---|
Temperature | BLR | Accuracy | TLU | 1.070 (not reported) | 0.035 | Larger herds associated with higher odds of correctly perceiving warming. |
MLR | Misjudgment | Age | 0.97 (0.96–0.98) | \(<0.001\) | Older respondents less likely to misjudge. | |
Education | 0.95 (0.93–0.97) | \(<0.001\) | Higher education reduces misjudgment. | |||
TLU | 0.93 (0.91–0.95) | \(<0.001\) | Livestock wealth decreases misjudgment odds. | |||
Underestimation | Education | 0.16 (0.14–0.18) | \(<0.001\) | Education strongly protects against underestimation. | ||
Cold spells | BLR | Accuracy | Age | 1.032 (not reported) | 0.027 | Older respondents are more likely to correctly perceive cold spells. |
Education | 1.129 (not reported) | 0.009 | Higher education improves cold-spell accuracy. | |||
Income | – | 0.017 | Higher income improves cold-spell accuracy. | |||
MLR | Overestimation | Income | 1.00 (1.0000–1.000002) | \(<0.001\) | Marginal effect: wealth linked to slight overestimation. | |
Rainfall | BLR | Accuracy | – | – | \(>0.10\) | No significant predictors found. |
MLR | Misjudgment | Age | 1.03 (1.01–1.05) | \(<0.001\) | Older respondents have slightly higher odds of misjudging rainfall. | |
Education | 0.85 (0.82–0.88) | \(<0.001\) | Higher education reduces rainfall misjudgment. | |||
Rain intensity | BLR | Accuracy (marginal) | Education | – | 0.068 | A marginal negative association between education and accuracy. |
MLR | Overestimation | Education | 1.08 (1.05–1.11) | \(<0.001\) | More education predicts overestimation. | |
Floods | BLR | Accuracy (marginal) | Age | – | 0.063 | A marginal positive association between age and accuracy. |
MLR | Overestimation | TLU | 1.02 (1.01–1.03) | \(<0.001\) | Larger herds linked to a slight increase in flood overestimation. | |
Warm spells | BLR | Accuracy | – | – | \(>0.10\) | No significant predictors found. |
Drought spells | BLR | Accuracy | – | – | \(>0.10\) | No significant predictors found. |
