But seriously, folks......
I have a windowless, unheated hangar. If I'm in the hangar and the engine is stone cold, and compare the OAT readout on my JPI probe with the readout on my "meat thermometer" probe, the temperature values shown will be identical. If I am parked outside, with the same sun shining on both probes at the same time (both probes being situated on the same pilot's side of the plane) again with the engine stone cold, the temperature readouts will once again be identical.
But if I start the engine, warm it up and go flying, the JPI (which has its probe just an inch behind the firewall) will show an OAT reading that is 2 degrees C warmer than my "meat thermometer" OAT thermometer which is mounted in the lower front corner of the pilot's window. There's nothing wrong with the JPI reading. It is giving a perfectly accurate reading of the ambient temperature where it happens to be situated. In other words, the engine is throwing off a lot of heat. I should have known better than to have mounted it right behind the engine compartment (and in my case, about 3 inches behind the big single muffler) of my 250.
I like Dave Gitelman's idea of mounting the probe in the wheel well. In fact, I wish I had mounted it almost anywhere else. One of these years, I'm going to paint the plane and when I do I'm going to move it.
Don Ostergard