Jason Samenow of the Washington Post reports that the National Weather Service has replaced its “temperature sensor” (i.e., thermometer) at Reagan National Airport. The previous sensor had been criticized as producing readings that were upwardly biased.
Regardless of one’s views about the science of climate change or what ought to be done about it, obtaining unbiased temperature data ought to be something everyone endorses.
For scientists and nonscientists alike, the problem of measurement error is a good reminder that data are not fixed in time and space. Rather, they are always affected by variability and uncertainty. That means when we observe differences between data points, whether is is differences in temperatures or called balls and strikes, those differences may be not be real.