Figure 5-3 shows the percentage of miles driven that included a maximum lateral or longitudinal
acceleration of a given degree, during each of the five phases of the experiment. Panel A shows the
case where the driver is alone in the vehicle. Panel B shows the case where there were one or more
passengers present. Note that the rates of these excessive maneuvers were generally more than
doubled when the driver was alone. This presumably is due to the lower threshold for a warning
when passengers are present, but could also be related to passenger presence itself or to trip
characteristics under which passenger trips occur. For the driver alone case, Panel A shows a drop in
the frequency of excessive maneuvers from the baseline to the treatment conditions; however, the
main effect of stage was not statistically significant (Table 5-3). Panel B also shows a drop from the
baseline to treatment conditions. The main effect with passengers present was statistically significant
only for 4-5 m/s2 case.
Analyses on the EMFS data indicated a main effect for time of day for the 2-3 m/s2 longitudinal
negative (e.g., slowing) dependent variable. Teen drivers reduced excessive maneuvers per mile at
nighttime compared to daytime. The results may also indicate a changing influence of traffic patterns
between the day and nighttime conditions. Additional braking, due to increased traffic in daytime
conditions could be influencing the results. As seen in Figure 5-4, there was also a general trend
observed for daytime excessive maneuvers to decrease over time during STC use and that it
remained lower than the baseline stage after STC use was discontinued for both nighttime and
daytime. These results support the contention that the EMFS does influence teen driver behavior by
reducing the number of excessive braking instances. The adaptive features of the subsystem are also
effective for excessive maneuvers because thresholds for this system became more restrictive at
night.
A significant main effect for stage was found for the 4-5m/s2 lateral negative maneuvers (i.e., right
turn) dependent variable for the passenger present dataset (see Figure 5-5). Post hoc analyses
indicated a significant reduction in these lateral maneuvers between the baseline and immediate
stages for this maneuver. Interestingly, there was a marked trend for the rate of lateral maneuvers
within the 3-4, and 4-5 m/s
2 dependent variables to decrease when teens were first exposed to the
STC and then remain lower than baseline levels throughout the remainder of the study. However,
the rate of lateral negative maneuvers for the 2-3 m/s2 dependent variable began to increase
throughout STC use but declined further and carried this lower effect into the transfer stage.
Collectively, the significant differences and the trends suggest a positive impact on teen driving
behaviors for lateral negative maneuvers that are experienced by teens on a more regular basis and
little effect for extreme maneuvers that are rarely experience (and thus situations in which feedback
is rarely provided).