Ideally, the assessment purposes should guide specific decisions we make during the construction or selection of an assessment tool, and ultimately the inferences we make with the assessment results. Early decisions that would be influenced by the decision-making context are factors such as the content of a test, the format of the tasks or items, the type of scaling, the degree of precision we need for particular score ranges, and the amount of error that we can tolerate in the results. We would then need to gather psychometric evidence to verify that the assessment tool generates information that is of adequate quality to support our inferences and specific decisions (this work is called validation). This chapter provided some examples of different types of validity evidence that might be needed for assessment tools used for different purposes. For example, predictive validity, reliability, or the dependability of cut-scores serves as important supporting evidence in assessments used for placement decisions (User Path 5). In sum, both assessment design and validation should be dictated by the needs of a specific situation. Later chapters in this book/module present a process model to help integrate the work related to assessment design/selection, validation, and use, keeping in mind the assessment purposes in given User Paths, the population, and the construct we want to measure.

Page 9