A Method to Assess Individualized Driver Models: Descriptiveness, Identifiability and Realism

•A novel model assessment method is proposed to assess driver model capabilities of capturing individual steering behaviour for practical identification purposes.•The model assessment method is demonstrated on two driver steering models, one of which has previously been used for individualized driver identification.•Key findings are that optimising and identification in purely the steering angle domain are insufficient and can be misleading.•Model assessment criteria provide an in-depth analysis why a model has satisfactory performance for the purpose of online identification of individualized driving behaviour.•Model assessment criteria give a quantitative grade on the models that also allows for comparison with other models.