When we use the full factorial model for any design for formulating any dosages form, we get responses that are significant and some response aren't. Since they have some values I would like to know aren't they affecting final result and why do we only consider the significant models in design study.