1.   In the first laboratory study subjective risk was only associated with improved recognition sensitivity for the most potentially dangerous situations.

2.   It thus appears that previous knowledge of a junction had almost no overall effect on recognition sensitivity in this task.

3.   Once again there is no hint of an overall inverted-U relationship or indeed any overall relationship between subjective risk and recognition sensitivity.

4.   This bias appears to be an effect which arises at time of retrieval and is independent of recognition sensitivity, Locksley et al.

5.   Therefore risk experienced at risky junctions may improve recognition sensitivity.

6.   Thus at non-risky junctions feelings of risk may actually impair recognition sensitivity by increasing the number of false alarms made.

7.   If there is little peripheral information in the non-risky exemplars, attention focusing would have no general effect on recognition sensitivity.

8.   Three subsequent laboratory studies, however, have failed to show any overall improvement in recognition sensitivity for films which were given high risk ratings.

9.   Specifically, those situations in which drivers might normally expect to experience risk showed the expected effect, recognition sensitivity was better for risky exemplars than less risky ones.