An automobile company analyzes warranty claim data reported by its dealer network to determine potential design problems with its vehicles. (a) Can warranty data be viewed as a random sample of failure? (b) Is this an observational study or a retrospective study?