In scientific publications, I like to refer to "misspecification(s)" of models. While the dictionary of my LaTeX-Editor complains that this is not a correct word, it is for example used in the wikipedia article on statistical model specification (https://en.wikipedia.org/wiki/Statistical_model_specification).
I am convinced that everyone reading my publicatons will understand what I mean by the word. So, is it a real word I can use, or should it be avoided? What would be a proper alternative?