Rose gets stuck on statistical significance
Too many takedowns to count for David Rose who doesn’t realize that short-term fluctuations in temperature tell you little about long term trends. The latest case is that according to one computer model, the temperature sequence ending in 2012 is close to the bottom edge of the statistical uncertainty range, a point where there’s only supposed to be a 5% chance that random variation produces a temperature below the modeled range.