I'm really not sure what this survey accomplishes. For a survey to any remote scientifically valid conclusion, one must survey total miles driven on every brand of tire being considered. In other words, the only valid statistic would be a failure per mile for X brand of tire. ANY OTHER CONCLUSION IS COMPLETELY INVALID. If 80% of OEM tires are of one particular brand, the reported failures will be disproportional for that brand. You need to have a separate survey for everyone who has not had a blowout to report the particular brand and how many blowout free miles they have driven.