Online Program Home
  My Program

Abstract Details

Activity Number: 517 - Issues in Transportation Statistics
Type: Contributed
Date/Time: Wednesday, August 2, 2017 : 10:30 AM to 12:20 PM
Sponsor: Transportation Statistics Interest Group
Abstract #322753 View Presentation
Title: Generalized Random-Effects Linear Models to Examine Driver Safety
Author(s): Ning Li* and Linda Ng Boyle
Companies: University of Washington and University of Washington
Keywords: driver safety ; general linear mixed models ; detection response task ; human factors ; driver distraction
Abstract:

The effects of cognitive demand from distracting tasks while driving can be assessed using a detection response task (DRT). A driving simulator study with 24 subjects was conducted to examine the sensitivity of DRT responses under 3 levels of task modes (auditory, visual, and a hybrid of both auditory and visual) and 2 levels of difficulty. Human factors studies contain limited sample sizes, which may impact the ability to reduce individual variance. Extremely skewed response data distributions and missing data can be problems that impact the relationship between the dependent and independent variables. Generalized linear mixed model are recommended to examine the association between the valid response time and the task conditions. The model was controlled for age group, gender, task modes, difficulty levels, and the subject random effect. The hypothesis is that increasing difficulty level is associated with increasing cognitive demand. Our findings suggest that response time is sensitive to workload changes among auditory, visual and hybrid tasks. These insights can only be obtained when we account for randomness and prior knowledge about participants.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2017 program

 
 
Copyright © American Statistical Association