That's not what he said at all
"" There is more to it, of course. If we're looking at the differences between the two groups, we can calculate a confidence interval (excel doesn't do this, so I plugged it somewhere else) - in this case, the 95% confidence interval for the mean drop in temps...