There is ALWAYS risk to be considered; it's engineering's job to IDENTIFY risks of a particular design, and suggest risk abatement, but MANAGEMENT must comprehend this and provide direction.
Not just identify, but quantify. And asses the impact of the abatement. When it comes down to basics, the only common denominator for risk/impact assessment is cost. Something Boeing was adverse to let engineers deal with*.
The other factor is understanding the fundamentals of the risk. In the words of Donald Rumsfelt, there are known unknowns and there are unknown unknowns. Boeing is adverse to admitting when they don't understand something. Their position being that they (and not the FAA/NTSB) are the sole authorities on the technologies, nobody else is in a position to second guess their choices. This is why we don't get an, "Oh crap! We should have used NiCads."
*In this case, I have to give McDonnell Douglas kudos. When Stonecipher came on board, he pushed for more cost visibility on the part of engineering. Heritage Boeing people pushed back.