Conference Proceeding
Authors: Alina Sinelnikova (Massachusetts Institute of Technology (AgeLab), Cambridge, MA) , Joonbum Lee (Massachusetts Institute of Technology (AgeLab), Cambridge, MA) , Bryan Reimer (Massachusetts Institute of Technology (AgeLab), Cambridge, MA) , Bruce Mehler (Massachusetts Institute of Technology (AgeLab), Cambridge, MA) , Joseph F Coughlin (Massachusetts Institute of Technology (AgeLab), Cambridge, MA)
This study examined differences in the impact of visual-manual and auditory-vocal based radio tuning tasks on field driving performance. Engagement in visual-manual tuning tasks were associated with higher steering wheel reversal rates than baseline driving. Both visual-manual and auditory-vocal based tuning tasks were associated with higher variances in speed maintenance compared to baseline driving. Models were built to utilize driving performance measurements as input to a classifier that aimed to distinguish between the three states (i.e., baseline driving, visual-manual tuning, and auditory-vocal tuning). Baseline driving could be classified from visual-manual tuning at an accuracy of over 99% and from auditory-vocal based tuning at an accuracy of 93.3%. Models could differentiate between the modalities at an accuracy of 75.2 % and between the three classes at an accuracy of 81.2%. Results suggest that changes in driving performance associated with visual-manual based tuning are statistically distinguishable from auditory-vocal based tuning. While not being free of visualmanual demand, tasks that involve auditory-vocal interactions appear to differ from visual-manual in how they impact driving performance.
Keywords:
How to Cite: Sinelnikova, A. , Lee, J. , Reimer, B. , Mehler, B. & Coughlin, J. (2015) “Predicting Secondary Task Involvement and Differences in Task Modality Using Field Highway Driving Data”, Driving Assessment Conference. 8(2015). doi: https://doi.org/10.17077/drivingassessment.1599