DescriptionIn the United States, the National Reporting System (NRS) serves as the accountability system for the federally funded adult education and literacy program. The NRS approves assessments which are used to show educational gain. Among other criteria, a test must demonstrate that it can reliably classify adult education students according to six Educational Functioning Levels (NRS, 2007). Only a handful of tests have been approved by the NRS for English language proficiency. One such test is the BEST Plus Oral English Proficiency Test (Best Plus). A study was designed to compare the suitability of the Versant English Test (VET) for use in classifying adult education students into the different NRS levels with that of the BEST Plus. Participants of 26 different language backgrounds representing the target population (n=151) successfully completed four counter-balanced tests for this study: two BEST Plus interviews and two Versant English Tests.
The data showed both tests to be internally reliable (r = .90 to .94) and fair indicators of each other (r = .74 to .78). However, an analysis of test scores revealed that it was difficult for either measure to consistently classify individuals according to NRS levels. A qualitative review of the scale descriptors reveals that the NRS bands represent such narrow gains in language proficiency that they are likely to overlap with measurement error in even the most reliable standardized tests. The implications for accountability of educational gain will be discussed.