Lane changing behavior is one of the fundamental driving behaviors seen in heterogeneous traffic conditions. The present study aims to analyze and model the path changing behavior of drivers, which is the microscopic version of lane changing behavior, at signalized intersections under the influence of access flows. For this, video-graphic survey was done for data collection from five four-legged intersection approaches in Kerala, followed by data extraction and analysis. Factors influencing the path change were identified by performing descriptive and inferential statistical analysis after the quantification of path change and included in the model. The choice of path change was modelled using binary logistic regression with clearance available, access type of vehicle and through type of vehicle as independent variables. The outcome of this study has potential application in microsimulation, intersection safety improvement and infrastructure planning.