The problems with testing a microwave oven's shieding effectiveness using a cell phone are legion:<p>1. The frequencies used by cell phones (typically 0.8 GHz in the U.S.) aren't the same as those used by an oven's magnetron (2.45 GHz), so we're comparing apples and oranges. The only valid test of shielding must necessarily use the same frequency as the microwave's magnetron.<p>2. The signal attenuation required to render a cell phone inoperative is not nearly enough to assure a person's safety when a magnetron is in operation. A power attenuation of 20 db (meaning reduction by a factor of 100) will almost certainly render a cell phone inoperative, but cannot assure safety to a person standing near a typical consumer microwave oven generating 0.7 kilowatts (700 watts) of microwave power and therefore emitting 7 watts into its immediate environment.<p>I emphasize the second example is hypothetical, meant only to show that an attenuation more than adequate to prevent the cell phone from working, is not nearly enough to assure a person's safety, not that this is a typical microwave oven leakage power.