The present disclosure relates to a method for assessing how a particular patient's responses may vary from those that are typically given on an item or instrument that relies on high granularity, dichotomous, responses. This method allows for the quantifiable description of patterns of possible overreporting, underreporting, “yeasaying,” “naysaying,” favoring one half of a representative continuum in a manner unrelated to item content, attempting to portray oneself in an overly favorable light, or attempting to portray oneself in an overly negative light. This method allows for validity checks without recourse to generating additional, specific, validity check items possibly modifying the times required to construct and train various machine learning (ML) systems including artificial neural networks (ANNs).
展开▼
