Abstract
This study was made to determine those items, or activities, of greatest relevance to motor-vehicle death rates. Specifi- cally, the purpose was 1) to find which of some 218 program and non-program items had statistically significant associa- tions with motor-vehicle death rates ( multiple regression analysis was used to determine these items). and 2) to study ways of refinding and improving the methods of evaluating activities which are intended to contribute to accident pre- vention ( for this, a new type of contingency table analysis was used in conjunction with the multiple regression analysis and a new method of measuring motor-vehicle deaths was employed).