Trini talk or the Queen's English? : navigating language varieties in the post-colonial, high stakes climate of "Standard Five" classrooms in Trinidad



Journal Title

Journal ISSN

Volume Title



This study was an exploration of the relationship between classroom discourse and a high-stakes, standardized test, the Secondary Entrance Assessment (SEA), administered to students in Trinidad in their final elementary school year ("Standard 5"), with scores used to determine placement into secondary school. Classroom discourse was measured as represented in the oral and written language modeled to students by teachers and textbooks, and compared to the written language expectations implicit in the instructions and items on the SEA. Students' SEA scores were analyzed in relation to their teachers' language to determine if achievement was related to exposure to local/non-standard features of Trinidadian English. One outcome of this analysis was the creation of a survey developed to measure teachers' propensity for speaking Trini (Creole/dialect) or Trinidadian Academic English (standard English) in class by having them respond to audio clips. The survey was found to have acceptable reliability and concurrent validity. For the oral language investigation, 13 teachers were recorded as they led their class in one lesson, and an index of Trini usage (number of Trini utterances per 100 words) was calculated for each teacher. This index was used in an HLM model to determine if teachers' language was a predictor of their students' SEA scores. The distribution of the Trini index was positively skewed, M=1.1 and SD=1.25, indicating low usage of Trini. Trini usage was not found to be a significant predictor of students' 2008 SEA scores, but was found to be a significant predictor of the variation in SEA scores when these were aggregated by class. Similarly, there were few instances of Trini features in the examined textbooks. However, on the SEA, the only influence of Trini was the presence of dialect options used as distractors on two "fill in the blank" items, suggesting a mismatch between the features used in instructional language and the language expectations on the exam. In general, the low exposure to local features did not account for students' achievement on the SEA, but was positively related to the consistency of scores within a class.