Predicting Proficiency without Direct Assessment: Can Speaking Ratings be Inferred from Non-participatory Listening and Reading Ratings.

摘要

This presentation, which was delivered to the Interagency Language Roundtable (ILR) Testing Committee on 22 MAR 2013, summarized four studies that provided evidence on the potential interchangeability of Defense Language Proficiency Test (DLPT) scores and Oral Proficiency Interview (OPI) ratings. The four studies addressed Briner and Rousseau s (2011) evidence-based approach to organizational decision-making. The first study was a meta-analytic review of the relationships among speaking, listening, and reading proficiency. Results indicated that, in general, listening and reading proficiency tests are not good proxies for speaking proficiency tests. The second and third studies explored whether DLPT listening (L) and reading (R) scores can be used as a proxy for determining OPI speaking proficiency ratings using two different samples. Findings indicated that neither the DLPT-L nor the DLPT-R are reliable or accurate predictors of OPI-S ratings. The fourth study explored Special Operations Forces (SOF) operators perceptions of the DLPT and the OPI. Specifically, the study explored whether SOF operators perceive that either test provides job-relevant and accurate assessments of language proficiency. Results indicated that SOF operators perceive that OPI is more related to job performance than the DLPT. Overall, these studies provide evidence that the OPI should remain the test of record for SOF.