The impact of immediate or delayed feedback on driving behaviour in a simulated Pay-As-You-Drive system

Chris Dijksterhuis*, Ben Lewis Evans, Bart Jelijs, Dick de Waard, Karel Brookhuis, Oliver Tucha

*Bijbehorende auteur voor dit werk

OnderzoeksoutputAcademicpeer review

19 Citaten (Scopus)


Pay-As-You-Drive (PAYD) insurance links an individual's driving behaviour to the insurance fee that they pay, making car insurance more actuarially accurate. The best known PAYD insurance format is purely mileage based and is estimated to reduce accidents by about 15% (Litman, 2011). However, these benefits could be further enhanced by incorporating a wider range of driving behaviours, such as lateral and longitudinal accelerations and speeding behaviour, thereby stimulating not only a safe but also an eco-friendly driving style. Currently, feedback on rewards and driver behaviour is mostly provided through a web-based interface, which is presented temporally separated from driving. However, providing immediate feedback within the vehicle itself could elicit more effect. To investigate this hypothesis, two groups of 20 participants drove with a behavioural based PAYD system in a driving simulator and were provided with either delayed feedback through a website, or immediate feedback through an in-car interface, allowing them to earn up to 6 extra. To be clear, every participant in the web group did actually view their feedback during the one week between sessions. Results indicate clear driving behaviour improvements for both PAYD groups as compared to baseline rides and an equal sized control group. After both PAYD groups had received feedback, the initial advantage of the in-car group was reduced substantially. Taken together with usability ratings and driving behaviours in specific situations these results show a moderate advantage of using immediate in-car feedback. However, the study also showed that under conditions of feedback certainty, the effectiveness of delayed feedback approaches that of immediate feedback as compared to a naive control group. (C) 2014 Elsevier Ltd. All rights reserved.

Originele taal-2English
Pagina's (van-tot)93-104
Aantal pagina's12
TijdschriftAccident Analysis and Prevention
Vroegere onlinedatum26-nov-2014
StatusPublished - feb-2015

Citeer dit