Not all data are created equal. Some data are good and some are bad. Even some good data – data collected using properly functioning equipment, unbiased, etc. – ought not to be used.
PBS.org has a good article on exactly this – data that ought not to be used – with respect to the recent high temps that most of the country is facing. Were 3,215 heat records broken? Or were 743? Or less? Or more? Or does the # of heat records broken even mean anything useful? Is it nothing more than an eye-catching headline?