Our Stout expert issued an opinion in a recall matter involving component failures in automotive aftermarket products installed in vehicles in multiple countries around the globe. Our analysis evaluated the results of sampling of several thousand installed products that had been removed from the field and tested for component failure. Our expert evaluated the stratified random sampling program and test-failure results to opine on whether there was a statistically sufficient and appropriate basis to conclude as to failures of the population of all installed products. Testifying at an international tribunal, our expert’s analysis considered test-failure results across multiple variables – production lot, country of installation, product version/release, product age, and the like. This analysis demonstrated that failure rates were correlated to premature aging of the component products, which were supposed to have a 10-year useful life but were typically failing within 3-4 years. Our expert’s analysis withstood critique by an opposing expert, a Ph.D. statistician and Fellow of the American Statistical Association, who suggested that an alternative hypothesis test would be more appropriate, and questioned the randomness and stratification of the sample. However, the opposing expert was found to have employed previously a virtually identical statistical methodology and approach as our expert’s.