Abstract
Using National Accident Sampling System files for the calendar years 1982-1986, shoulder belt use and effectiveness were compared among 1974-1980 model year cars with and without window shade devices, which human factors research has found to produce excess slack in the belts. Although shoulder belt use was not correlated to presence or absence of the window shade mechanism when model year and calender year effects were controlled statistically, the effectiveness of belts in cars with window shade device was consistently less than in cars without the device, particularly so in larger vehicles.