Change in the percentage of cyclists wearing helmets per unit change in explanatory variables
| Model with external variables only | Model including individual characteristics | |||
|---|---|---|---|---|
| Coefficient (95% CI) | p Value | Coefficient (95% CI) | p Value | |
| Predicted high temperature (°F) | 0.018 (−0.101 to 0.137) | 0.765 | 0.011 (−0.095 to 0.118) | 0.832 |
| Rain chance | ||||
| No chance | Ref. | Ref. | Ref. | Ref. |
| Under 50% | −2.126 (−5.420 to 1.169) | 0.203 | −1.685 (−4.458 to 1.087) | 0.230 |
| 50% or more | −1.232 (−8.806 to 6.342) | 0.747 | −0.211 (−7.125 to 6.702) | 0.952 |
| Actual rain | 2.54 (−7.531 to 12.620) | 0.616 | 0.385 (−10.313 to 11.084) | 0.943 |
| Evening | −10.589 (−14.274 to −6.904) | <0.001 | −7.917 (−12.158 to −3.676) | <0.001 |
| Friday | −5.204 (−9.236 to −1.172) | 0.012 | −2.436 (−6.567 to 1.694) | 0.243 |
| Bikeshare riders (Δ%) | – | – | −0.441 (−0.680 to −0.202) | <0.001 |
| Compression Pants wearers (Δ%) | – | – | 0.324 (0.015 to 0.633) | 0.040 |
The model also controls for observation site (not significant). CIs and p values use Newey-West SEs, which are robust SEs further adjusted for possible serial correlation. In these models, the coefficient represents the change in the percentage of cyclists wearing helmets per unit change in each explanatory variable .