Are you serious? It's pretty easy to argue with findings when his sample is limited to 25 routes and 1 set of dates for each route. It may be indicative of a larger trend, but it certainly not statistically significant.