While high-stakes testing may return similar outcomes in the tests of different students, which is an element of reliability, one might wonder whether it is reliability that America really needs. In this regard, it is necessary to examine the validity, where this paper proposes that validity is more important as far as this testing approach is concerned.
Validity has been defined as the extent to which a test measures what it is meant to measure. If a test is valid, it measures exactly what it is meant to measure as purely as not to, by chance, influence any other factors. With validity, focus is not precisely on the scores measured, but on the inferences that one is able to deduce from the instruments. As such, the inferences made from a valid test are supposed to be “suitable, meaningful, and useful” (Lang & Wilkerson, 2008). This is the complex link that sheds clean light on the clear distinction between validity and reliability. In other words, as one would infer, it is possible for a testing instrument to measure something apart from the construct that it was supposed to measure, and do it reliably.
On the other hand, a measure that is not reliable can never be said to be valid. In this light, reliability is quite necessary a measure, but it is quite inadequate in relation to validity – a valid instrument, therefore, has to be reliable, which a reliable instrument does not have to be valid. In testing, it is a common understanding that based on the aforementioned relationship, violations of validity are expected to have more severe impacts, as compared to reliability. One would, therefore, agree that validity is more complex and important an instrument that reliability and is less understood compared to the latter. In addition, it cannot be substantiated by any one statistic. Having mentioned that, it is important to discuss the important