One is the set of tests performed by the U.K. CAA some years ago, in which they measured the field strength of cell phone transmissions in various parts of an airplane, and found that in the cockpit area at least the field strength in some locations was stronger than previously supposed. The report is on the CAA WWW site.
Unfortunately it's not but it does appear to be here:
Interference Levels In Aircraft at
Radio Frequencies used by Portable Telephones
They tested various powers, frequencies and aircraft and concluded that a 2W transmission (maximum a phone radiates) from the forward cabin in a 737-236 could produce a 4.5 V/m field in the flight deck. The highest seen in the avionics bay was 1.87 V/m (747-243).
The report further quotes the RF immunity standards for avionics, in which it's quite clear that equipment approved prior to December 1989 might not reach a sufficient standard of protection (0.1 V/m test level), but equipment approved after that date P, Y, W, V, U and T require immunity to at least 5 V/m.
The CAA also tested the effect of RF on various bits of avionics:
Effects of Interference from Cellular Telephones on Aircraft Avionic Equipment
In that test, they picked equipment certified to the pre-1989 0.1 V/m standard and exposed it to 25 to 50 V/m. In an experiment with results about as shocking as Newton's apple falling
downwards from the tree, they discovered that sometimes the avionics didn't work properly
-- though in fact below 30 V/m they saw almost no effect.
This is a difficult topic, because relatively small failure rates (e.g. 1 per million flights) might be deemed unacceptable, so limited lab tests cannot necessarily provide useful data. But from the CAA papers, I certainly drew the conclusion that GSM phones were extremely unlikely to have an effect on avionics that justified the current prohibition and its resultant costs on business and quality of life.