Suppose, just for argument’s sake, that when your son Johnny turned 2 years old you decided to monitor his growth. His birthday is Jan. 1, so on the first of every month you measure his height — you even mount a tape measure permanently on the wall so you can measure him in the same location each time. You dutifully record the number for each measurement. Your wife thinks this is “cute” so she decides to take a picture every time you do, and puts them in a photo album labelled “Our Growing Child.” Since she’s a professional photographer, she uses her amazing Nikon super-high-res digital camera. All her friends think it’s super-cute.
Come October of your child’s 3rd year, you plot a graph of his height as a function of time:
“Oh my God!” you exclaim to your wife. “Our child is shrinking! Call the doctor! Take him to the emergency room! Quick!!!”
Your wife looks at the graph and says, “That can’t be right. He didn’t shrink by 5 cm in the last month — we’d have noticed. Besides, look here in my photo album. You can see that the final number isn’t right. You must have written it down wrong — it’s simply an error.”
But you’re still worried. So you consult your next-door neighbor, who happens to work with the Global Warming Policy Foundation, and show the data to him. He declares, “Your wife must have some pro-growth political agenda. The data clearly show that your child is shrinking. You’d better take him to the doctor immediately.”
Your wife staunchly refuses to give in to your irrationality. But the next day your neighbor knocks on your door and says, “I can prove that your child is either shrinking, or at least has stopped growing. Look at this graph of his height since age 3 years, 2 months. Over the last 8 months your child has shrunk at a rate of 1.3 cm/yr.”
Your wife interrupts, saying “Please don’t be so ridiculous. Obviously the last data point is in error, this photograph proves it.” The neighbor retorts, “Even if I leave out the last data point, the recent growth rate is not statistically significant! And the estimated rate is only 1.5 cm/yr, far less than the expected rate of about 7 cm/yr according to the IPCC.” (Intergovernmental Panel on Child Care)
You call the doctor, frantically, and make an appointment for first thing in the morning next day. Your wife insists on talking to him, saying she thinks you’re mistaken and that she will investigate in detail.
She looks over her photo album. That’s when she notices that sometimes when you measured little Johnny’s height, he had his shoes on. Other times he was barefoot. Fortunately, since she photographed all the measurements she knows when he was barefoot and when he was shod. She carefully measures the thickness of the soles of his shoes at 2 cm. This, she says, is an exogenous factor which has nothing to do with his actual growth, it just causes random fluctuations in the measurements.
She also enlarges her latest photo (she has all the digital images) and is able to read the correct height from the image — those Nikon cameras are awesome!
She corrects the final data point. Then she subtracts 2 cm from each measurement for which little Johnny had his shoes on. This gives her an adjusted data set, which she states is a much better, much more correct, representation of Johnny’s growth. She also logs on the internet and retrieves standard growth data from the World Health Organization. Finally, she compares the adjusted data to the WHO standard:
Your next-door neighbor refuses to believe it. He says your wife faked the data, that she’s trying to “hide the decline.”
Meanwhile, your wife calls the doctor and cancels the appointment. Smart woman.