Abstract
A FIELD SURVEY WAS CONDUCTED TO EVALUATE TWO PROCEDURES DESIGNED TO MEASURE THE EFFECTIVENESS OF MOTOR VEHICLE BRAKING SYSTEMS. SELECTED FAILURE RATES AND AGREEMENT MEASURES WERE COMPUTED USING A RECENTLY DEVELOPED UNIFIED APPROACH TO THE ANALYSIS OF MULTIVARIATE CATEGORICAL DATA. IT WAS FOUND THAT THE PROCEDURES AGREE ONLY WEAKLY, AND THAT THE AGREEMENT VARIED WITH CERTAIN PASS/FAIL CRITERIA. ON THE BASIS OF CONDITIONAL ARGUMENTS, A MOVING-STOPPING TEST WAS FOUND TO BE MORE STRINGENT THAN A WHEEL REMOVING INSPECTION.(Author/publisher).