[klausnrooster]

That's all true. I am biased because I am responsible for keeping hundreds of devices calibrated, and keeping a lab within 2C of 23C. The CLT implies that the sample mean is more likely to be close to the true population mean as the sample size increases. Not the same as more accurate. Shame on me!

[random314]

The number of people in this thread not understanding the difference between bias and variance errors is too much.

http://scott.fortmann-roe.com/docs/BiasVariance.html