They are awful. With that sample size, the accuracy is barely enough to reliably call a 55-45% split, which nowadays is also frequently labeled a 'landslide'. For the recent closely contested elections, the metapolls, or poll of polls with combined sample size approaching 10,000, have been much better. The 2000 U.S. Presidential election was so close that a sample size exceeding 100,000 would have been needed to reliably call the national plurality, which didn't match the electoral contest.
I bought my Prius on July 17, and my dealer only had 2 left. The one I test drove was already reserved for a buyer. The dealer said they were going by quickly.
OK, I was just trying to point out that when you have a large population to sample, it is not the sample size relative to the poplulation size that matters, its the absolute sample size that matters. The statistical fluctuations depend on the sample size, not the sample percentage relative to the general population you are sampling. Therefore, knowing the total number of Prius sold is not really the issue. Do you agree?
Yes, I agree. My dispute was really off topic, with the opinion poll sample size needed to predict national elections. Two or three decades ago, these polls normally questioned 1200-1500 people, and were fairly good. In the past decade, at the same time many contests became closer, many pollsters saved money by shrinking the sample sizes. When questioned why different pollsters get different results, and why the daily tracking polls are so erratic, they give many excuses, but always avoid mentioning Problem #1 - Sample Size. Their samples are simply too small to meaningfully call close contests.
Agree. Yes, that is unbelievable. My dealer will NOT do that at all. If they are spoken for, no one touches it. Our car only had 1 mile on it when we picked it up.
I'm not sure if that is common practice or not. I didn't know until after I test drove it that it was reserved for another customer. But I had no interest in getting it anyway because I hated the color.