And they didn't even realize the actual conditions. The ice team had taken temperature readings that showed the booster well below ambient temperature, these were dismissed as wrong. They weren't wrong, the weather that morning was calm enough the LOX boiloff was falling pretty close to straight down, cooling the booster.
The tragedy of Challenger is that Morton Thiokol engineers did look enough to know that the O-ring design was likely to fail under cold weather conditions, and they specifically recommended that Challenger postpone launch because of the temperature on the pad the night before the launch.
They got overruled for political reasons (the big bosses didn't want to throw off the timing of the first class being taught in space).
I don't think that the "teacher in space" part was much of a factor in Marshall's decision-making. There was clearly plan-continuation bias, and there was a clear directive that Marshall should not be responsible for delaying a launch. The management style they had inherited from von Braun also meant that disagreement was highly discouraged. The data was ambiguous enough that they could convince themselves that everything was fine and Morton Thiokol was coming up with a new launch requirement at the last minute.
There's a really well written and researched book on the topic, that I highly recommend, called Truth, Lies, and O-Rings by Allan J. McDonald, James R. Hansen, and Jonathan Yen
The Challenger disaster could perfectly well have been avoided if they had been willing to look. There was a clear indication of substantial deviation from design and the joint could have been engineered to stay in compression during the twang.
But Columbia was fundamental. There was an awful lot of insulation aerodynamically before the heat shield. Once again, multiple warnings, but keeping the insulation from being torn off while keeping the weight down was problematic. Columbia was the beginning of the end as they realized they couldn't sweep it under the rug anymore and they couldn't truly fix it.