Our team has developed a statistical procedure (1992) that relates every answer to every other answer without regard to whether it is "right" or "wrong."
Such relationships can inform teachers the nature as well as the depth of understanding held by students at the moment of their giving their answers.
We went of track when we made the erroneous assumption that "wrong" answers are "blind" guesses and not worthy, therefore, of attention.
With this erroneous assumption we have derailed our entire educational process because we have turned our attention to content replication accuracy instead of contextual validity.
It is long past time for a change. Research evidence supporting this claim has been in the literature for more than 30 years and the "uncertainty principle" for nearly 100 years.
Shame on us!