Validity refers to the accuracy of inferences we make from an assessment’s results. In theory, our inferences will be perfectly valid when all systematic differences in the responses of individuals we measure are caused by actual differences on the measured construct, and not by any other external interference (systematic error). Cultural biases in the questions that might give a particular group of respondents an advantage/disadvantage when taking a test is an example of a systematic error that would affect validity. We determine the validity of inferences by verifying the extent to which we are successful in keeping systematic, external inferences out of an assessment tool and its results. This work is called validation.

Page 3