Double loop learning occurs when mismatches or errors [difference,
d, between actual and desired states] are corrected by first
examining and altering the governing variables and then taking action.
Governing variables are the preferred states that individuals strive to
"satisfice" when they are acting. These governing variables are not
the underlying beliefs or values people espouse. They are the
variables that can be inferred, by observing the actions of
individuals acting as agents for the organisation, to drive and
guide their actions.