The real mistake happened around 1976 when they settled on the basic design which they estimated had a 2-3% chance of blowing up on any launch.
Every other manned space vehicle has had an emergency escape for the crew. An emergency escape system could have saved the lives of the Challenger crew because they survived the external tank explosion (were killed when the crew compartment hit the ocean) -- they thought about putting one in, but then they wouldn't have had space in the hatchback for a satellite.
The thermal tiles were also "unsafe at any speed", it was known at the beginning that the tiles would get broken or fall off sometime and there was a lot of nail-biting around the first flights, but after "it" didn't happen a few times people started to relax.
Sociologist Dianne Vaughn popularized the term "Normalization of Deviance" in connection with the Challenger Disaster but it has been greatly misunderstood with the public.
Today people think it is the doctor not washing his hands and not being confronted by the nurse about it, but in the case of NASA it is a formal process where they make a list of 200 unacceptable situations about the space shuttle that they make the case they can squeak by one by one and each speaker has a few minutes to make a case and if they had bad slides and didn't make a good point they would clear the floor for the next "catastrophe".
"Normalization of Deviance" is a regular procedure with dangerous technology -- it's the paperwork that they fill out because they put the door frame in the wrong way in the break room at the nuclear powerplant or when they fly a regional jet back from a regional airport with one angle-of-attack sensor down because it can be fixed much more easily at the big airport.