The effect of a program to stop drunk driving on how often people drive depends on the program's aims, how it is put into action, how data is collected, how statistics are analyzed, how control groups are used, how behavioral surveys are done, and how the results are put together.Why justify this hypothesis is a good statement?