English: 2000 Car Brand Quality by Infrequency of Trouble-Prone Models

To form a brand-quality measure from the 2000 list of Used Cars To Avoid, the first step is to count each brand's entries on the list. Each model year of each model is treated as a separate entry.

Next, as the number of automobile models sold under a brand name varies greatly from brand to brand, it is necessary to take account of the fact that a brand with more models has a greater opportunity to have more model years of low quality. To compensate for a possibly inflated, or deflated, frequency of trouble-prone model years within a brand, as well as a variability in model data sufficiency, the number of a brand's entries in CR's 2000 Used-Cars-To-Avoid list is divided by the total number of overall reliability ratings for the brand found in the reliability charts of the same issue of Consumer Reports. The overall reliability ratings are found in the Reliability-Verdict row of the 2000 reliability charts.

By the method of computation, this quality measure begins with 0 and may run to a value some greater than 1. The value of 0 is the highest quality rating attainable by a brand and is achieved only when a brand has no entry on the Used-Cars-To-Avoid list.

The quality ranking of the car brands by the foregoing computations, together with their quality ratings, are given in the bar graph above. Only those brands with at least 5 overall CR reliability ratings are included.

For a Google Knol that summarizes the changes in car brand and manufacturer ranking by this quality measure from 1990 to 2010, visit http://knol.google.com/k/james-bleeker/automobile-quality-by-car-brand-and/3o3u27f0whd6k/2?hd=ns# .