A puzzled economist phoned an acoustician friend of his with a surprising question: Could a door without doorstep and seals reach a sound reduction index value of 38 dB? Assured that there was no chance of that, he blurted out that he had been given a test report as proof! The so-called test report was a simple sound reduction index curve without any trace of leakage. The acoustician insisted that a full test report should include the description of the sample, and the full report was eventually sent to the economist. On examining the full document, it turned out that the contractor had simply tested the door panel without its frame.
In the course of a large performance facility project, one of the roof manufacturers in competition boasted of better acoustic performances than competitors and had a laboratory test report to show it. Looking closely at the description it read, “The panels are laid on the mineral wool”; it turned out that many mechanical fixations had actually been omitted compared to the normal mounting procedure (and such a fixation would never have been accepted by the safety engineer).
A large round-robin test was held in the 1980s  on a basic plasterboard partition. Testing the supposedly same kind of plasterboard partition (i.e., made of materials from the same origin), the deviation was ±6 dB on the global value. The spread of values in the 125 Hz band could reach 25 dB.