Comparing Crowdsourced Near-Miss and Collision Cycling Data and Official Bike Safety Reporting
Bibliographic record
Abstract
Official sources of cyclist safety data suffer from underreporting and bias. Crowdsourced safety data have the potential to supplement official sources and to provide new data on near-miss incidents. BikeMaps.org is a global online mapping tool that allows cyclists to record the location and details of near misses and collisions they experience. However, little is known about how the characteristics of near-miss and collision events compare. Further, the question remains whether the characteristics of crowdsourced collision data are similar to those of collision data captured by official insurance reports. The objectives of this study were twofold: ( a) to assess similarities and differences in near misses and collisions reported to BikeMaps.org and ( b) to assess similarities and differences in collisions reported to BikeMaps.org and to an official insurance data set. Logistic regression was used first to model the odds of crowdsourced near-miss reports as opposed to collision reports and then to model the odds of crowdsourced as opposed to official insurance collision reports, as a function of incident circumstances. The results indicated higher odds of crowdsourced reports of near misses than of crowdsourced collision reports for commute trips, interactions with motor vehicles, and in locations without bicycle-specific facilities. In addition, relative to insurance reports, crowdsourced collision reports were associated with peak traffic hours, nonintersection locations, and locations where bicycle facilities were present. These analyses indicated that crowdsourced collision data have potential to fill in gaps in reports to official collision sources and that crowdsourced near-miss reporting may be influenced by perceptions of risk.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.020 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.008 | 0.003 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.000 |
| Research integrity | 0.000 | 0.002 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; both teacher heads agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".