Just a bit of silliness for the end of a Friday: Velvet Spicer at the Rochester Business Journal reports that consumer confidence in Rochester is skyrocketing!!! Why it rose a whole point from last month:
Overall confidence in upstate—which includes current and future confidence—was 53.4 in November, compared with 49.5 in October and 67.7 a year ago.
You have to ask yourself: in a pool as small as respondents to a poll concerning Upstate, how small does a change need to be before it’s disregarded as statistical noise? In other words, is a one percent change in such a small pool of data a genuine change or just a reflection of the variations that are normal with an opinion poll?
Think about it: if they asked 800 people what they thought of the economy, one percent of respondents would only be eight people! How is this statistically significant in any way?