An article entitled "Hurricane Love Songs" reports that Hurricane Charley's winds in August 2004 "apparently didn't faze fish in the midst of spawning. University of South Florida oceanographers, who regularly listen to the 'love songs' fish sing while spawning, said they expected Charley's winds and the Increased water flow from its rain might silence the chorus. But when they reviewed tapes made from their hydrophones when Charley hit Charlotte Harbor, they found the fish just sang louder."18 The oceanographers could have begun by setting up a null hypothesis that the spawning songs were equally loud before and during the hurricane, against the alternative that they were less loud during the hurricane. In a case like this where the data actually contradict the onesided alternative hypothesis, what can you say about the size of the P-value?