She measured sound levels at the ear, not in the airflow. She specifically mentions children being shorter and likely to stand closer to the noise source, and so made readings at multiple distances.