On evaluating monitoring design effectiveness