11.   Above this point there is a dramatic increase in the probability of recall with increasing risk ratings.

12.   The interaction between risk rating and accident estimate across the ten junctions suggests that subjects were able to sensibly dissociate the two judgment tasks in the laboratory setting.

13.   However, the magnitude and direction of the correlations appears to be related to the overall mean risk rating for the junction.

14.   The most interesting results of the study were the relationships observed between risk ratings and P within these sets of items.

15.   It appeared that for exemplars of the generally risky junctions risk ratings were positively correlated with P.

16.   Within the six exemplars of each junction correlations of P with the risk ratings from the previous study were calculated.

17.   In fact there is no sign of any overall relationship between risk ratings and P.

18.   Because there is no overall relationship between risk ratings and recognition performance it is necessary to consider the effects that may operate within individual junctions.

19.   These numbers are surprisingly low, no junction had a mean accident or risk rating above the midpoint of the scale.

20.   The mean risk rating given in the Groeger and Chapman study was in fact considerably higher than that given by subjects in this experiment.