The document summarizes research analyzing factors that predict wrong-way biking violations at a university campus. Data was collected over 14 hours and analyzed using logistic regression. The final model identified gender, having a bag while wearing sportswear, having a bag with higher speed, and biking within 5 minutes of class ending times as significant predictors. The model achieved an accuracy of 81.3% but suffered from low R2 values and significant overlap between predicted probabilities for violators and non-violators. Other classifiers like Parzen windows achieved slightly better accuracy.