Descriptive and normative ethical behavior appear to be functionally distinct

Cox (2021) ran two quick experiments with college students. First, each student read short stories about tough choices. They picked the 'right' action and wrote why they picked it.

Next, the same students saw new stories. This time the wording or the outcome changed a little. The team watched whether the person's choice, or their reason, shifted with the new details.

Most people did not line up. Their pick of 'right' action and their reason for it often moved in different directions.

When the context changed, sometimes only the choice flipped. Other times only the reason flipped. The two parts acted like separate behaviors, not one glued unit.

Bahry et al. (2023) extend the same idea to autism services. They say we must check both 'Is this goal ethical?' and 'Will it help long-term life?' Cox shows those questions can be scored on their own.

Perez et al. (2021) found that a simple color cue can switch what an equivalent picture means—happy, scary, or nothing. Cox found a parallel: a small story tweak can switch ethical 'what' or 'why' without touching the other.

Lyon (1982) warned that two behaviors seen together still need separate tests. Cox supplies the test: measure choice and justification apart, then change context to see which one moves.

Next time you review a parent goal or staff policy, ask two questions and record two answers. First, 'Do you agree this is right?' Second, 'Why do you say it's right?' Change the context—add a new consequence or stakeholder—and ask again. If only one answer shifts, you now have data showing the choice and the reason sit under different environmental controls. That tells you which lever to adjust in training or policy, instead of assuming they move together.