Empirical Studies on Traffic Flow in Inclement Weather
Figure 5.1. This figure presents an overview of the statistical analysis performed on the data. The Weather Adjustment Factors were utilized a stepwise regression analysis, from which a set of optimum models was generated. Models were validated for significance (with a p value). If they were not significant, the set of models was revised. If the model significance is satisfactory, variable significance is evaluated. If a variable is not, it is removed from the model and the model is re-evaluated for significance. If all variables in the model are significant, outlier data is removed from the data set. If the r-squared value is below 20 percent, the model equals constant (average factor) and the final model is established. If it is not, a residual normality test is performed. If it fails the test, rank data and re-perform stepwise regression analysis. If the residual normality test is satisfactory, a variance test is performed. If the variance test fails, transform the data and re-perform the stepwise regression analysis. If the variance test is satisfactory, proceed to the final model.
