This study aimed to investigate the utility of the responses of the two child dummies (P1.5 and P3) that are placed in the rear seat, in identical forwardfacing child restraints during frontal Australian NCAP (ANCAP) tests.

Dynamic responses of the two child dummies, vehicle crash parameters, and frontal dummy responses were extracted from the ANCAP report database for 35 frontal crash tests. Linear regression analysis was used to assess:​ the similarity between the two dummies’ responses;​ variation between frontal dummy responses;​ and relationships between the child dummy responses and other measured crash parameters.

Dynamic responses from the P1.5 and P3 dummies were highly correlated with each other, including head accelerations, neck forces, and chest accelerations (p<0.0001 for all, 0.4 < R2 < 0.6). Variation between the two rear-seated child dummies was substantially less than between the driver and front passenger dummies. The child dummies’ head and chest accelerations were correlated to vehicle b-pillar deceleration (p≤0.01 for all), but not to vehicle mass, vehicle class, or other crash parameters (p>0.05 for all).

Unlike the two front-seated occupants, where the dummies provide different information about the vehicle’s safety performance, the two rear-seated child dummies in child restraints are providing essentially duplicate information. Head excursion of the dummies is not measured in the current ANCAP test protocol, and this may be a more sensitive and meaningful assessment of child restraint occupant serious head injury risk. Only 35 vehicles were included in the analysis, and data on some variables (including neck moments, and harness and top tether payout during testing) were not recorded in all tests.

These results suggest that using two child dummies in forward-facing child restraints is not providing significantly more information than could be gleaned from a single child dummy in a forwardfacing child restraint. This suggests that one of these child dummies could be usefully replaced with an alternative dummy representing an older rear seat occupant, without loss of information on a vehicle’s ability to protect child-restraint users. Possibilities for such a replacement occupant include a 10 year old child dummy using the lapsash seatbelt (as is being trialed in Japan NCAP tests), a booster-seated 6 year old dummy, or a small female occupant. Any of these options would provide additional information on vehicle safety performance than is currently being reported in most NCAPs.