In rural areas 25 percent, and in urban areas over 50 percent, of accidents in sweden occur at junctions. Road authorities need information to: 1) determine the `normal' number of accidents, injury consequences and conflict type distributions; 2) estimate traffic safety effects of countermeasures. There are methodological difficulties in making these studies. Cross section studies normally underestimate effects of countermeasures while before and after studies normally overestimate them. A method based on the empirical bayes method is proposed. The accident rate, under comparable traffic conditions, is 1.5-2 times higher for 4-way than for 3-way junctions. It may be advantageous to replace 4-way junctions by staggered 3-way junctions. Lighting is effective means of reducing accidents. Traffic islands on secondary roads reduce accidents at 4-way junctions by 4 percent. Kerbstone channelisation of left-turn lanes on primary road is of doubtful value in rural areas. Channelisation by painting only is effective at rural 3-way junctions, reducing accidents by about 20 percent. It is also cheaper. Urban 4-way junctions can be improved by roundabouts or traffic lights. Roundabouts reduce injury consequences of accidents by 50 percent while traffic lights cut the number of accidents by up to 50 percent. Lowering of speed limit at junctions significantly reduces severity of accidents. For comments on this paper, see C 6529 and C 6530.
Abstract