Questioning the currency of language certification exams as measures of 21st century competencies.
Implicit in the results of all standardized language exams, such as those bearing certification, is a professional guarantee, attested to by some reputable institution, that the scores from the assessment represent an individual’s specific level of knowledge or proficiency in some domain of language use (e.g., CEFR), such that the scores can ultimately be used to predict future performance. Resulting scores are then used to grant rewards to or impose sanctions on individuals–all in the service of public policy goals. Given the potentially deleterious consequences of misclassifications, however, the claims made by such high-stakes exams require empirical justification by a rigorous program of validation aligned with established professional standards—i.e., the Standards for Educational and Psychological Testing (AERA, APA, NCME, 2014) or the International Language Testing Association’s Guidelines for Practice (2010). Such accountably measures serve to bolster public trust, as the scores from these exams function as powerful symbols of fair, equitable, and just decision-making in a meritocracy.
Given the aspirational goal of predicting L2 performance with today’s real-world tasks, however, one must logically question the currency of an older exam, and wonder when it is time for any well-known, standardized assessment to reflect upon its currency, and to courageously engage in the complex task of refreshment—if, for no other than the simple reason that the kinds of language-driven real world tasks we were asked to do twenty years ago represent an unequivocally different skill set from that needed to function in today’s fast-moving, technology-driven, globalized world. For example, today’s educational institutions and its agile workforce obviously need individuals with foundational knowledge and relevant experience, but perhaps more than ever, it needs individuals able to use language and other resources to build, co-construct in teams, and share new understandings through reasoning and problems-solving skills. Society also needs individuals able to communicate across diverse groups and functional boundaries and in varying contexts, as well as individuals able to adapt, learn from feedback, and regulate not only their learning and performance, but also their motivational and attitudinal dispositions. Finally, we need professionals of all sorts to be able embrace the use of new information and communication technologies (Breu et al., 2001; OCED, 2001, Stasz, 2000). The question then is to what degree do the constructs in current certification exams reflect these modern-day, language driven competencies when making predictions about future L2 performance?
In this talk I will question the currency of several traditional certification exams and their ability to make predictions about language-related performance that reflect competencies currently valued in the real world. I will also discuss design decision choices that promote broadened rather than reduced test constructs. Finally, I will provide some thoughts for a path forward.