Urban Cyclist Survey Responses in Loja, Ecuador (2025)

Research Hypothesis:
Urban cyclists’ self-reported behaviors and demographics influence—but do not fully explain—the accuracy of their risk perception relative to objective infrastructure risk. Behavioral patterns, route familiarity, and cycling experience interact in complex ways, sometimes leading to paradoxical outcomes.

Data Collection and Structure:
This dataset contains responses from 153 urban cyclists in Loja, Ecuador, gathered between April and August 2025 via an online questionnaire and field collection with tablets. A non-probabilistic, convenience sampling method was used through local cycling groups and in-person recruitment.

The survey included four sections:
-Demographics: Age, sex, cycling frequency, primary purpose of cycling, and years of experience.
-Cycling Behavior (CBQ): Scores for violations, errors, and combined risky behaviors, plus positive behaviors.
-Perceived Risk Assessments: Ratings of 12 road segment photographs (Low, Medium, High, Extreme, coded 1–4) and frequency of use for each site.
-Infrastructure Perception: General assessment of cycling infrastructure quality and prioritized safety improvements (e.g., driver education, physical separation, better signage).

Data Interpretation and Use:
The dataset supports analysis of the misalignment between subjective and objective risk and identification of cyclist profiles with high or low perception accuracy. Key findings include:
-Behavioral paradox: Frequent and experienced cyclists with high CBQ violation/error scores showed the poorest risk perception.
-Alertness effect: Low-frequency users were occasionally more accurate on medium-risk segments.
-Systematic misjudgment: High and extreme risks were underestimated, while low risks were overestimated.
-Infrastructure perceptions: Most participants rated the network as “Regular” or worse, prioritizing driver education and separated lanes for safety.

Limitations:
-Male-dominated sample (~84%), reflecting local cycling demographics but limiting generalizability.
-Static photographs cannot replicate the dynamic, multisensory nature of on-road cycling.
-Self-reported behaviors are subject to recall and social desirability biases.

This dataset is valuable for urban safety studies, behavioral analysis, and designing interventions to align cyclist perception with actual risk, complementing infrastructure-based risk assessments.