Validityis perceivable as a comprehensive, evaluative result of the level towhich visible evidence and theoretical justifications support theappropriateness and adequacy of actions and interpretations usingmetrics such as test scores (Messick,1995, p. 741).Validity, therefore, is not the product of an assessment or a test,but the meaning that is deduced from the trial scores. In thisregard, the interpretation or meaning of the score is what theresearchers consider valid, in addition to the implications for theactions that this definition encompasses. The common description ofvalidity separates into three categories: construct, content, andcriterion validities. This model has been proven to be incomplete andfragmented because it does not consider the substantiation of thevalue suggestions of the score meaning – as a starting point foraction – and the associated outcomes of score use. Unified validitybrings together the social values and score meaning in testutilization and interpretation, which makes it the ideal tool foreducational investigations.
Unifiedvalidity puts together the reflections advanced by criteria, content,and consequence into a construct structure. As a result, an idealenvironment for conducting the pragmatic testing of reasonablehypotheses about theoretically harmonious relationships and scoremeaning is developed (Messick,1995, p. 741).Construct validity has six distinct aspects that have been forwardedas a way of addressing the core issues that are implicit in the ideaof validity being viewed as a unified theory. The facets includestructural, substantive, generalizability, consequential, content,and external components. In operation, these six features work as thegeneral validity criteria, and, are also perceivable as the standardsfor all psychological and educational measurement and performanceassessments.
Consideringthe above assertions, unified validity is ideal for educationalresearch for various reasons. For instance, student portfolios areusually sources of deduction – about the essential quality ofeducational products and the skills, knowledge, and other traits ofstudents – and inferences regarding the quality and constructsrequired to realize the standards of validity (Messick,1995, p. 741).This allegation is premised on the fact that performance assessmentshave been forwarded as the instruments of standard-based educationreform because they have been associated with positive consequencesfor learning and teaching. Such outcomes necessitate theimplementation of a systematic approach to validity.
Themethodical approach to validity also works for crucial measurementissues like comparability, reliability, and fairness. The fairnessreference brings to bear a comprehensive set of equity concepts intesting. Topics such as freedom from bias in interpretation andscoring, the evenhandedness of test use, and the suitability of therules that underlie decision-making and test-based constructs emergedistinctly. Thus, the concepts reliability, validity, fairness, andcomparability do not just measure principles they also have ameaning for the forces and implications that stretch beyondmeasurement whenever evaluative decisions and judgments are made(Messick,1995, p. 742).As a significant social value, validity encompasses both a politicaland scientific function. This role cannot be fulfilled by acorrelation coefficient between a purported criterion and a testscore or expert judgments that check whether content and the proposedtest use are in tandem.
Ina recap of the above discussion, unified validity is an importanttool in educational investigations because it brings together socialvalues and scores meaning in test utilization and interpretation.This validity type is essential in educational research because itmerges the manifestations advanced by criteria, content, andconsequence into a construct arrangement. As a result, the perfectenvironment for carrying out the pragmatic testing of reasonablehypotheses about theoretically pertinent relationships and scoremeaning is developed, as discussed above.
Messick,S. (1995). Validity of Psychological Assessment. AmericanPsychological Association,50(9),741 – 749.