Among other poor statistical methodology, it's not as though the average respondent to this poll is likely to have a reasonable sample size of appliances to com

Among other poor statistical methodology, it's not as though the average respondent to this poll is likely to have a reasonable sample size of appliances to compare. They're not saying "Out of the six refrigerators I own, the Kenmore breaks down most often." It's actually "I own a Kenmore refrigerator and it broke down at some point.", which is more or less entirely meaningless. If they sampled the population to find the proportion of owners of each brand that reported their appliance breaking and compared those results it would be slightly more meaningful.